435 U.S. 519 (1978)2
Supreme Court of United States.
Argued November 28, 1977.
Decided April 3, 1978.
CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE DISTRICT OF COLUMBIA CIRCUIT.5
 Thomas G. Dignan, Jr., argued the cause for petitioner in No. 76-419. With him on the briefs were G. Marshall Moriarty, William L. Patton, and R. K. Gad III. Charles A. Horsky argued the cause for petitioner in No. 76-528. With him on the briefs was Harold F. Reis.6
Deputy Solicitor General Wallace argued the cause for the federal respondents in support of petitioners in both cases pursuant to this Court's Rule 21 (4). On the briefs were Solicitor General McCree, Acting Assistant Attorney General Liotta, Harriet S. Shapiro, Edmund B. Clark, John J. Zimmerman, Peter L. Strauss, and Stephen F. Eilperin. Henry V. Nickel and George C. Freeman, Jr., filed a brief for respondents Baltimore Gas & Electric Co. et al. in support of petitioner in No. 76-419 pursuant to Rule 21 (4).7
Richard E. Ayres argued the cause and filed briefs for respondents in No. 76-419. Myron M. Cherry argued the cause for the nonfederal respondents in No. 76-528. With him on the brief was Peter A. Flynn.8
In 1946, Congress enacted the Administrative Procedure Act, which as we have noted elsewhere was not only "a new, basic and comprehensive regulation of procedures in many agencies," Wong Yang Sung v. McGrath, 339 U. S. 33 (1950), but was also a legislative enactment which settled "long-continued and hard-fought contentions, and enacts a formula upon which opposing social and political forces have come to rest." Id., at 40. Section 4 of the Act, 5 U. S. C. § 553 (1976 ed.), dealing with rulemaking, requires in subsection (b) that  "notice of proposed rule making shall be published in the Federal Register . . . ," describes the contents of that notice, and goes on to require in subsection (c) that after the notice the agency "shall give interested persons an opportunity to participate in the rule making through submission of written data, views, or arguments with or without opportunity for oral presentation. After consideration of the relevant matter presented, the agency shall incorporate in the rules adopted a concise general statement of their basis and purpose." Interpreting this provision of the Act in United States v. Allegheny-Ludlum Steel Corp., 406 U. S. 742 (1972), and United States v. Florida East Coast R. Co., 410 U. S. 224 (1973), we held that generally speaking this section of the Act established the maximum procedural requirements which Congress was willing to have the courts impose upon agencies in conducting rulemaking procedures. Agencies are free to grant additional procedural rights in the exercise of their discretion, but reviewing courts are generally not free to impose them if the agencies have not chosen to grant them. This is not to say necessarily that there are no circumstances which would ever justify a court in overturning agency action because of a failure to employ procedures beyond those required by the statute. But such circumstances, if they exist, are extremely rare.10
Even apart from the Administrative Procedure Act this Court has for more than four decades emphasized that the formulation of procedures was basically to be left within the discretion of the agencies to which Congress had confided the responsibility for substantive judgments. In FCC v. Schreiber, 381 U. S. 279, 290 (1965), the Court explicated  this principle, describing it as "an outgrowth of the congressional determination that administrative agencies and administrators will be familiar with the industries which they regulate and will be in a better position than federal courts or Congress itself to design procedural rules adapted to the peculiarities of the industry and the tasks of the agency involved." The Court there relied on its earlier case of FCC v. Pottsville Broadcasting Co., 309 U. S. 134, 138 (1940), where it had stated that a provision dealing with the conduct of business by the Federal Communications Commission delegated to the Commission the power to resolve "subordinate questions of procedure . . . [such as] the scope of the inquiry, whether applications should be heard contemporaneously or successively, whether parties should be allowed to intervene in one another's proceedings, and similar questions."11
It is in the light of this background of statutory and decisional law that we granted certiorari to review two judgments of the Court of Appeals for the District of Columbia Circuit because of our concern that they had seriously misread or misapplied this statutory and decisional law cautioning reviewing courts against engrafting their own notions of proper procedures upon agencies entrusted with substantive functions by Congress. 429 U. S. 1090 (1977). We conclude that the Court of Appeals has done just that in these cases, and we therefore remand them to it for further proceedings. We also find it necessary to examine the Court of Appeals' decision with respect to agency action taken after full adjudicatory hearings. We again conclude that the court improperly intruded into the agency's decisionmaking process, making it necessary for us to reverse and remand with respect to this part of the cases also.12
Under the Atomic Energy Act of 1954, 68 Stat. 919, as amended, 42 U. S. C. § 2011 et seq., the Atomic Energy Commission  was given broad regulatory authority over the development of nuclear energy. Under the terms of the Act, a utility seeking to construct and operate a nuclear power plant must obtain a separate permit or license at both the construction and the operation stage of the project. See 42 U. S. C. §§ 2133, 2232, 2235, 2239. In order to obtain the construction permit, the utility must file a preliminary safety analysis report, an environmental report, and certain information regarding the antitrust implications of the proposed project. See 10 CFR §§ 2.101, 50.30 (f), 50.33a, 50.34 (a) (1977). This application then undergoes exhaustive review by the Commission's staff and by the Advisory Committee on Reactor Safeguards (ACRS), a group of distinguished experts in the field of atomic energy. Both groups submit to the Commission their own evaluations, which then become part of the record of the utility's application. See 42 U. S. C. §§ 2039, 2232 (b). The Commission staff also undertakes the review required by the National Environmental Policy Act of 1969 (NEPA), 83 Stat. 852, 42 U. S. C. § 4321 et seq., and prepares a draft environmental impact statement, which, after being circulated for comment, 10 CFR §§ 51.22-51.25 (1977), is revised and becomes a final environmental impact statement. § 51.26. Thereupon a three-member Atomic Safety and Licensing Board conducts a public adjudicatory hearing, 42 U. S. C. § 2241, and reaches a decision which can be  appealed to the Atomic Safety and Licensing Appeal Board, and currently, in the Commission's discretion, to the Commission itself. 10 CFR §§ 2.714, 2.721, 2.786, 2.787 (1977). The final agency decision may be appealed to the courts of appeals. 42 U. S. C. § 2239; 28 U. S. C. § 2342. The same sort of process occurs when the utility applies for a license to operate the plant, 10 CFR § 50.34 (b) (1977), except that a hearing need only be held in contested cases and may be limited to the matters in controversy. See 42 U. S. C. § 2239 (a); 10 CFR § 2.105 (1977); 10 CFR pt. 2, App. A, V (f) (1977).15
These cases arise from two separate decisions of the Court of Appeals for the District of Columbia Circuit. In the first, the court remanded a decision of the Commission to grant a license to petitioner Vermont Yankee Nuclear Power Corp. to operate a nuclear power plant. Natural Resources Defense Council v. NRC, 178 U. S. App. D. C. 336, 547 F. 2d 633 (1976). In the second, the court remanded a decision of that same agency to grant a permit to petitioner Consumers Power Co. to construct two pressurized water nuclear reactors to generate electricity and steam. Aeschliman v. NRC, 178 U. S. App. D. C. 325, 547 F. 2d 622 (1976).16
In December 1967, after the mandatory adjudicatory hearing and necessary review, the Commission granted petitioner Vermont Yankee a permit to build a nuclear power plant in Vernon, Vt. See 4 A. E. C. 36 (1967). Thereafter, Vermont Yankee applied for an operating license. Respondent Natural Resources Defense Council (NRDC) objected to the granting  of a license, however, and therefore a hearing on the application commenced on August 10, 1971. Excluded from consideration at the hearings, over NRDC's objection, was the issue of the environmental effects of operations to reprocess fuel or dispose of wastes resulting from the reprocessing operations. This ruling was affirmed by the Appeal Board in June 1972.18
In November 1972, however, the Commission, making specific reference to the Appeal Board's decision with respect to the Vermont Yankee license, instituted rulemaking proceedings "that would specifically deal with the question of consideration of environmental effects associated with the uranium fuel cycle in the individual cost-benefit analyses for light water cooled nuclear power reactors." App. 352. The notice of proposed rulemaking offered two alternatives, both predicated on a report prepared by the Commission's staff entitled Environmental Survey of the Nuclear Fuel Cycle. The first would have required no quantitative evaluation of the environmental hazards of fuel reprocessing or disposal because the Environmental Survey had found them to be slight. The second would have specified numerical values for the environmental impact of this part of the fuel cycle, which values would then be incorporated into a table, along with the other relevant factors, to determine the overall cost-benefit balance for each operating license. See id., at 356-357.19
Much of the controversy in this case revolves around the  procedures used in the rulemaking hearing which commenced in February 1973. In a supplemental notice of hearing the Commission indicated that while discovery or cross-examination would not be utilized, the Environmental Survey would be available to the public before the hearing along with the extensive background documents cited therein. All participants would be given a reasonable opportunity to present their position and could be represented by counsel if they so desired. Written and, time permitting, oral statements would be received and incorporated into the record. All persons giving oral statements would be subject to questioning by the Commission. At the conclusion of the hearing, a transcript would be made available to the public and the record would remain open for 30 days to allow the filing of supplemental written statements. See generally id., at 361-363. More than 40 individuals and organizations representing a wide variety of interests submitted written comments. On January 17, 1973, the Licensing Board held a planning session to schedule the appearance of witnesses and to discuss methods for compiling a record. The hearing was held on February 1 and 2, with participation by a number of groups, including the Commission's staff, the United States Environmental Protection Agency, a manufacturer of reactor equipment, a trade association from the nuclear industry, a group of electric utility companies, and a group called Consolidated National Intervenors which represented 79 groups and individuals including respondent NRDC.20
After the hearing, the Commission's staff filed a supplemental document for the purpose of clarifying and revising the Environmental Survey. Then the Licensing Board forwarded its report to the Commission without rendering any decision. The Licensing Board identified as the principal procedural question the propriety of declining to use full formal adjudicatory procedures. The major substantive issue was the technical adequacy of the Environmental Survey.21
 In April 1974, the Commission issued a rule which adopted the second of the two proposed alternatives described above. The Commission also approved the procedures used at the hearing, and indicated that the record, including the Environmental Survey, provided an "adequate data base for the regulation adopted." Id., at 392. Finally, the Commission ruled that to the extent the rule differed from the Appeal Board decisions in Vermont Yankee "those decisions have no further presidential significance," id., at 386, but that since "the environmental effects of the uranium fuel cycle have been shown to be relatively insignificant, . . . it is unnecessary to apply the amendment to applicant's environmental reports submitted prior to its effective date or to Final Environmental Statements for which Draft Environmental Statements have been circulated for comment prior to the effective date," id., at 395.22
Respondents appealed from both the Commission's adoption of the rule and its decision to grant Vermont Yankee's license to the Court of Appeals for the District of Columbia Circuit.23
In January 1969, petitioner Consumers Power Co. applied for a permit to construct two nuclear reactors in Midland,  Mich. Consumers Power's application was examined by the Commission's staff and the ACRS. The ACRS issued reports which discussed specific problems and recommended solutions. It also made reference to "other problems" of a more generic nature and suggested that efforts should be made to resolve them with respect to these as well as all other projects. Two groups, one called Saginaw and another called Mapleton, intervened and opposed the application. Saginaw filed with the Board a number of environmental contentions, directed over 300 interrogatories to the ACRS, attempted to depose the chairman of the ACRS, and requested discovery of various ACRS documents. The Licensing Board denied the various discovery requests directed to the ACRS. Hearings were then held on numerous radiological health and safety issues. Thereafter, the Commission's staff issued a draft  environmental impact statement. Saginaw submitted 119 environmental contentions which were both comments on the proposed draft statement and a statement of Saginaw's position in the upcoming hearings. The staff revised the statement and issued a final environmental statement in March 1972. Further hearings were then conducted during May and June 1972. Saginaw, however, choosing not to appear at or participate in these latter hearings, indicated that it had "no conventional findings of fact to set forth" and had not "chosen to search the record and respond to this proceeding by submitting citations of matters which we believe were proved or disproved." See App. 190 n. 9. But the Licensing Board, recognizing its obligations to "independently consider the final balance among conflicting environmental factors in the record," nevertheless treated as contested those issues "as to which intervenors introduced affirmative evidence or engaged in substantial cross examination." Id., at 205, 191.25
At issue now are 17 of those 119 contentions which are claimed to raise questions of "energy conservation." The Licensing Board indicated that as far as appeared from the record, the demand for the plant was made up of normal industrial and residential use. Id., at 207. It went on to state that it was "beyond our province to inquire into whether the customary uses being made of electricity in our society are `proper' or `improper.'" Ibid. With respect to claims that Consumers Power stimulated demand by its advertising the Licensing Board indicated that "[n]o evidence was offered on this point and absent some evidence that Applicant is creating abnormal demand, the Board did not consider the  question." Id., at 207-208. The Licensing Board also failed to consider the environmental effects of fuel reprocessing or disposal of radioactive wastes. The Appeal Board ultimately affirmed the Licensing Board's grant of a construction permit and the Commission declined to further review the matter.26
At just about the same time, the Council on Environmental Quality revised its regulations governing the preparation of environmental impact statements. 38 Fed. Reg. 20550 (1973). The regulations mentioned for the first time the necessity of considering in impact statements energy conservation as one of the alternatives to a proposed project. The new guidelines were to apply only to final impact statements filed after January 28, 1974. Id., at 20557. Thereafter, on November 6, 1973, more than a year after the record had been closed in the Consumers Power case and while that case was pending before the Court of Appeals, the Commission ruled in another case that while its statutory power to compel conservation was not clear, it did not follow that all evidence of energy conservation issues should therefore be barred at the threshold. In re Niagara Mohawk Power Corp., 6 A. E. C. 995 (1973). Saginaw then moved the Commission to clarify its ruling and reopen the Consumers Power proceedings.27
In a lengthy opinion, the Commission declined to reopen the proceedings. The Commission first ruled it was required to consider only energy conservation alternatives which were "`reasonably available,'" would in their aggregate effect curtail demand for electricity to a level at which the proposed facility would not be needed, and were susceptible of a reasonable degree of proof. App. 332. It then determined, after a thorough examination of the record, that not all of Saginaw's contentions met these threshold tests. Id., at 334-340. It further determined that the Board had been willing at all times to take evidence on the other contentions. Saginaw had simply failed to present any such evidence. The  Commission further criticized Saginaw for its total disregard of even those minimal procedural formalities necessary to give the Board some idea of exactly what was at issue. The Commission emphasized that "[p]articularly in these circumstances, Saginaw's complaint that it was not granted a hearing on alleged energy conservation issues comes with ill grace." Id., at 342. And in response to Saginaw's contention that regardless of whether it properly raised the issues, the Licensing Board must consider all environmental issues, the Commission basically agreed, as did the Board itself, but further reasoned that the Board must have some workable procedural rules and these rules28
"in this setting must take into account that energy conservation is a novel and evolving concept. NEPA `does not require a "crystal ball" inquiry.' Natural Resources Defense Council v. Morton, [148 U. S. App. D. C. 5, 15, 458 F. 2d 827, 837 (1972)]. This consideration has led us to hold that we will not apply Niagara retroactively. As we gain experience on a case-by-case basis and hopefully, feasible energy conservation techniques emerge, the applicant, staff, and licensing boards will have obligations to develop an adequate record on these issues in appropriate cases, whether or not they are raised by intervenors.
"However, at this emergent stage of energy conservation principles, intervenors also have their responsibilities. They must state clear and reasonably specific energy conservation contentions in a timely fashion. Beyond that, they have a burden of coming forward with some  affirmative showing if they wish to have these novel contentions explored further." Id., at 344 (footnotes omitted).
Respondents then challenged the granting of the construction permit in the Court of Appeals for the District of Columbia Circuit.30
With respect to the challenge of Vermont Yankee's license, the court first ruled that in the absence of effective rulemaking proceedings, the Commission must deal with the environmental impact of fuel reprocessing and disposal in individual licensing proceedings. 178 U. S. App. D. C., at 344, 547 F. 2d, at 641. The court then examined the rulemaking proceedings and, despite the fact that it appeared that the agency employed all the procedures required by 5 U. S. C. § 553 (1976 ed.) and more, the court determined the proceedings to be inadequate and overturned the rule. Accordingly, the Commission's determination with respect to Vermont Yankee's license was also remanded for further proceedings. 178 U. S. App. D. C., at 358, 547 F. 2d, at 655.32
 With respect to the permit to Consumers Power, the court first held that the environmental impact statement for construction of the Midland reactors was fatally defective for  failure to examine energy conservation as an alternative to a plant of this size. 178 U. S. App. D. C., at 331, 547 F. 2d, at 628. The court also thought the report by ACRS was inadequate, although it did not agree that discovery from individual ACRS members was the proper way to obtain further explication of the report. Instead, the court held that the Commission should have sua sponte sent the report back to the ACRS for further elucidation of the "other problems" and their resolution. Id., at 335, 547 F. 2d, at 632. Finally, the court ruled that the fuel cycle issues in this case were controlled by NRDC v. NRC, discussed above, and remanded for appropriate consideration of waste disposal and other unaddressed fuel cycle issues as described in that opinion. 178 U. S. App. D. C., at 335, 547 F. 2d, at 632.33
Petitioner Vermont Yankee first argues that the Commission may grant a license to operate a nuclear reactor without any consideration of waste disposal and fuel reprocessing. We find, however, that this issue is no longer presented by the record in this case. The Commission does not contend that it is not required to consider the environmental impact of the spent fuel processes when licensing nuclear power plants. Indeed, the Commission has publicly stated subsequent to the Court of Appeals' decision in the instant case that consideration of the environmental impact of the back end of the fuel cycle in "the environmental impact statements for individual LWR's [light-water power reactors] would represent a full and candid assessment of costs and benefits consistent with the legal requirements and spirit of NEPA." 41 Fed. Reg. 45849 (1976). Even prior to the Court of Appeals' decision the Commission implicitly agreed that it would consider the back end of the fuel cycle in all licensing proceedings: It indicated that it was not necessary to reopen prior licensing proceedings because "the environmental effects of the uranium fuel cycle have been shown to be relatively insignificant," and thus incorporation of those effects into the cost-benefit analysis would not change the results of such licensing proceedings. App. 395. Thus, at this stage of the proceedings the only question presented for review in this regard is whether the Commission may consider the environmental impact of the fuel processes when licensing nuclear reactors. In addition to the weight which normally attaches to the agency's determination of such a question, other reasons support the Commission's conclusion.36
Vermont Yankee will produce annually well over 100 pounds of radioactive wastes, some of which will be highly toxic. The Commission itself, in a pamphlet published by its  information office, clearly recognizes that these wastes "pose the most severe potential health hazard . . . ." U. S. Atomic Energy Commission, Radioactive Wastes 12 (1965). Many of these substances must be isolated for anywhere from 600 to hundreds of thousands of years. It is hard to argue that these wastes do not constitute "adverse environmental effects which cannot be avoided should the proposal be implemented," or that by operating nuclear power plants we are not making "irreversible and irretrievable commitments of resources." 42 U. S. C. §§ 4332 (2) (C) (ii), (v). As the Court of Appeals recognized, the environmental impact of the radioactive wastes produced by a nuclear power plant is analytically indistinguishable from the environmental effects of "the stack gases produced by a coal-burning power plant." 178 U. S. App. D. C., at 341, 547 F. 2d, at 638. For these reasons we hold that the Commission acted well within its statutory authority when it considered the back end of the fuel cycle in individual licensing proceedings.37
We next turn to the invalidation of the fuel cycle rule. But before determining whether the Court of Appeals reached a permissible result, we must determine exactly what result it did reach, and in this case that is no mean feat. Vermont Yankee argues that the court invalidated the rule because of the inadequacy of the procedures employed in the proceedings. Brief for Petitioner in No. 76-419, pp. 30-38. Respondents, on the other hand, labeling petitioner's view of the decision a "straw man," argue to this Court that the court merely held that the record was inadequate to enable the reviewing court to determine whether the agency had fulfilled its statutory obligation. Brief for Respondents in No. 76-419, pp. 28-30, 40. But we unfortunately have not found the parties' characterization of the opinion to be entirely reliable; it appears here, as in Orloff v. Willoughby, 345 U. S. 83, 87 (1953), that  "in this Court the parties changed positions as nimbly as if dancing a quadrille."39
After a thorough examination of the opinion itself, we conclude  that while the matter is not entirely free from doubt, the majority of the Court of Appeals struck down the rule because of the perceived inadequacies of the procedures employed in the rulemaking proceedings. The court first determined the intervenors' primary argument to be "that the decision to preclude `discovery or cross-examination' denied them a meaningful opportunity to participate in the proceedings as guaranteed by due process." 178 U. S. App. D. C., at 346, 547 F. 2d, at 643. The court then went on to frame the issue for decision thus:40
"Thus, we are called upon to decide whether the procedures provided by the agency were sufficient to ventilate the issues." Ibid., 547 F. 2d, at 643.
The court conceded that absent extraordinary circumstances it is improper for a reviewing court to prescribe the procedural format an agency must follow, but it likewise clearly thought it entirely appropriate to "scrutinize the record as a whole to insure that genuine opportunities to participate in a meaningful way were provided . . . ." Id., at 347, 547 F. 2d, at 644. The court also refrained from actually ordering the agency to follow any specific procedures, id., at 356-357, 547 F. 2d, at 653-654, but there is little doubt in our minds that  the ineluctable mandate of the court's decision is that the procedures afforded during the hearings were inadequate. This conclusion is particularly buttressed by the fact that after the court examined the record, particularly the testimony of Dr. Pittman, and declared it insufficient, the court proceeded to discuss at some length the necessity for further procedural devices or a more "sensitive" application of those devices employed during the proceedings. Ibid., The exploration of the record and the statement regarding its insufficiency might initially lead one to conclude that the court was only examining the sufficiency of the evidence, but the remaining portions of the opinion dispel any doubt that this was certainly not the sole or even the principal basis of the decision. Accordingly, we feel compelled to address the opinion on its own terms, and we conclude that it was wrong.42
In prior opinions we have intimated that even in a rulemaking proceeding when an agency is making a "`quasijudicial'" determination by which a very small number of persons are "`exceptionally affected, in each case upon individual grounds,'" in some circumstances additional procedures may be required in order to afford the aggrieved individuals due process. United States v. Florida East Coast R. Co., 410 U. S., at 242, 245, quoting from Bi-Metallic Investment Co. v. State Board of Equalization, 239 U. S. 441, 446 (1915). It might also be true, although we do not think the issue is presented in this case and accordingly do not decide it, that a totally unjustified departure from well-settled agency procedures of long standing might require judicial correction.43
 But this much is absolutely clear. Absent constitutional constraints or extremely compelling circumstances the "administrative agencies `should be free to fashion their own rules of procedure and to pursue methods of inquiry capable of permitting them to discharge their multitudinous duties.'" FCC v. Schreiber, 381 U. S., at 290, quoting from FCC v. Pottsville  Broadcasting Co., 309 U. S., at 143. Indeed, our cases could hardly be more explicit in this regard. The Court has, as we noted in FCC v. Schreiber, supra, at 290, and n. 17, upheld this principle in a variety of applications, including that case where the District Court, instead of inquiring into the validity of the Federal Communications Commission's exercise of its rulemaking authority, devised procedures to be followed by the agency on the basis of its conception of how the public and private interest involved could best be served. Examining § 4 (j) of the Communications Act of 1934, the Court unanimously held that the Court of Appeals erred in upholding that action. And the basic reason for this decision was the Court of Appeals' serious departure from the very basic tenet of administrative law that agencies should be free to fashion their own rules of procedure.44
We have continually repeated this theme through the years, most recently in FPC v. Transcontinental Gas Pipe Line Corp., 423 U. S. 326 (1976), decided just two Terms ago. In that case, in determining the proper scope of judicial review of agency action under the Natural Gas Act, we held that while a court may have occasion to remand an agency decision because of the inadequacy of the record, the agency should normally be allowed to "exercise its administrative discretion in deciding how, in light of internal organization considerations, it may best proceed to develop the needed evidence and how its prior decision should be modified in light of such evidence as develops." Id., at 333. We went on to emphasize:45
"At least in the absence of substantial justification for doing otherwise, a reviewing court may not, after determining that additional evidence is requisite for adequate  review, proceed by dictating to the agency the methods, procedures, and time dimension of the needed inquiry and ordering the results to be reported to the court without opportunity for further consideration on the basis of the new evidence by the agency. Such a procedure clearly runs the risk of `propel[ling] the court into the domain which Congress has set aside exclusively for the administrative agency.' SEC v. Chenery Corp., 332 U. S. 194, 196 (1947)." Ibid.
Respondent NRDC argues that § 4 of the Administrative Procedure Act, 5 U. S. C. § 553 (1976 ed.), merely establishes lower procedural bounds and that a court may routinely require more than the minimum when an agency's proposed rule addresses complex or technical factual issues or "Issues of Great Public Import." Brief for Respondents in No. 76-419, p. 49. We have, however, previously shown that our decisions reject this view. Supra, at 542 to this page. We also think the legislative history, even the part which it cites, does not bear out its contention. The Senate Report explains what eventually became § 4 thus:47
"This subsection states . . . the minimum requirements of public rule making procedure short of statutory hearing. Under it agencies might in addition confer with industry advisory committees, consult organizations, hold informal `hearings,' and the like. Considerations of practicality, necessity, and public interest . . . will naturally govern the agency's determination of the extent to which public proceedings should go. Matters of great import, or those where the public submission of facts will be either useful to the agency or a protection to the public, should naturally be accorded more elaborate public procedures." S. Rep. No. 752, 79th Cong., 1st Sess., 14-15 (1945).
The House Report is in complete accord:49
"`[U]niformity has been found possible and desirable for all classes of both equity and law actions in the courts . . . .  It would seem to require no argument to demonstrate that the administrative agencies, exercising but a fraction of the judicial power may likewise operate under uniform rules of practice and procedure and that they may be required to remain within the terms of the law as to the exercise of both quasi-legislative and quasi-judicial power.'.....
"The bill is an outline of minimum essential rights and procedures. . . . It affords private parties a means of knowing what their rights are and how they may protect them . . . ......
". . . [The bill contains] the essentials of the different forms of administrative proceedings . . . ." H. R. Rep. No. 1980, 79th Cong., 2d Sess., 9, 16-17 (1946).
And the Attorney General's Manual on the Administrative Procedure Act 31, 35 (1947), a contemporaneous interpretation previously given some deference by this Court because of the role played by the Department of Justice in drafting the legislation, further confirms that view. In short, all of this leaves little doubt that Congress intended that the discretion of the agencies and not that of the courts be exercised in determining when extra procedural devices should be employed.51
There are compelling reasons for construing § 4 in this manner. In the first place, if courts continually review agency proceedings to determine whether the agency employed procedures which were, in the court's opinion, perfectly tailored to reach what the court perceives to be the "best" or "correct" result, judicial review would be totally unpredictable. And the agencies, operating under this vague injunction to employ  the "best" procedures and facing the threat of reversal if they did not, would undoubtedly adopt full adjudicatory procedures in every instance. Not only would this totally disrupt the statutory scheme, through which Congress enacted "a formula upon which opposing social and political forces have come to rest," Wong Yang Sung v. McGrath, 339 U. S., at 40, but all the inherent advantages of informal rulemaking would be totally lost.52
Secondly, it is obvious that the court in these cases reviewed the agency's choice of procedures on the basis of the record actually produced at the hearing, 178 U. S. App. D. C., at 347, 547 F. 2d, at 644, and not on the basis of the information available to the agency when it made the decision to structure the proceedings in a certain way. This sort of Monday morning quarterbacking not only encourages but almost compels the agency to conduct all rulemaking proceedings with the full panoply of procedural devices normally associated only with adjudicatory hearings.53
Finally, and perhaps most importantly, this sort of review fundamentally misconceives the nature of the standard for judicial review of an agency rule. The court below uncritically assumed that additional procedures will automatically result in a more adequate record because it will give interested parties more of an opportunity to participate in and contribute to the proceedings. But informal rulemaking need not be based solely on the transcript of a hearing held before an agency. Indeed, the agency need not even hold a formal hearing. See 5 U. S. C. § 553 (c) (1976 ed.). Thus, the adequacy of the "record" in this type of proceeding is not correlated directly to the type of procedural devices employed, but rather turns on whether the agency has followed the statutory mandate of the Administrative Procedure Act or other relevant statutes. If the agency is compelled to support  the rule which it ultimately adopts with the type of record produced only after a full adjudicatory hearing, it simply will have no choice but to conduct a full adjudicatory hearing prior to promulgating every rule. In sum, this sort of unwarranted judicial examination of perceived procedural shortcomings of a rulemaking proceeding can do nothing but seriously interfere with that process prescribed by Congress.54
Respondent NRDC also argues that the fact that the Commission's inquiry was undertaken in the context of NEPA somehow permits a court to require procedures beyond those specified in § 4 of the APA when investigating factual issues through rulemaking. The Court of Appeals was apparently also of this view, indicating that agencies may be required to "develop new procedures to accomplish the innovative task of implementing NEPA through rulemaking." 178 U. S. App. D. C., at 356, 547 F. 2d, at 653. But we search in vain for something in NEPA which would mandate such a result. We have before observed that "NEPA does not repeal by implication any other statute." Aberdeen & Rockfish R. Co. v. SCRAP, 422 U. S. 289, 319 (1975). See also United States v. SCRAP, 412 U. S. 669, 694 (1973). In fact, just two Terms ago, we emphasized that the only procedural requirements imposed by NEPA are those stated in the plain language of the Act. Kleppe v. Sierra Club, 427 U. S. 390, 405-406 (1976). Thus, it is clear NEPA cannot serve as the basis for a substantial revision of the carefully constructed procedural specifications of the APA.55
In short, nothing in the APA, NEPA, the circumstances of this case, the nature of the issues being considered, past agency practice, or the statutory mandate under which the Commission operates permitted the court to review and overturn the rulemaking proceeding on the basis of the procedural devices employed (or not employed) by the Commission so long as the Commission employed at least the statutory minima, a matter about which there is no doubt in this case.56
 There remains, of course, the question of whether the challenged rule finds sufficient justification in the administrative proceedings that it should be upheld by the reviewing court. Judge Tamm, concurring in the result reached by the majority of the Court of Appeals, thought that it did not. There are also intimations in the majority opinion which suggest that the judges who joined it likewise may have thought the administrative proceedings an insufficient basis upon which to predicate the rule in question. We accordingly remand so that the Court of Appeals may review the rule as the Administrative Procedure Act provides. We have made it abundantly clear before that when there is a contemporaneous explanation of the agency decision, the validity of that action must "stand or fall on the propriety of that finding, judged, of course, by the appropriate standard of review. If that finding is not sustainable on the administrative record made, then the Comptroller's decision must be vacated and the matter remanded to him for further consideration." Camp v. Pitts, 411 U. S. 138, 143 (1973). See also SEC v. Chenery Corp., 318 U. S. 80 (1943). The court should engage in this kind of review and not stray beyond the judicial province to explore the procedural format or to impose upon the agency its own notion of which procedures are "best" or most likely to further some vague, undefined public good.57
We now turn to the Court of Appeals' holding "that rejection of energy conservation on the basis of the `threshold test'  was capricious and arbitrary," 178 U. S. App. D. C., at 332, 547 F. 2d, at 629, and again conclude the court was wrong.60
The Court of Appeals ruled that the Commission's "threshold test" for the presentation of energy conservation contentions was inconsistent with NEPA's basic mandate to the Commission. Id., at 330, 547 F. 2d, at 627. The Commission, the court reasoned, is something more than an umpire who sits back and resolves adversary contentions at the hearing stage. Ibid., 547 F. 2d, at 627. And when an intervenor's comments "bring `sufficient attention to the issue to stimulate the Commission's consideration of it,'" the Commission must "undertake its own preliminary investigation of the proffered alternative sufficient to reach a rational judgment whether it is worthy of detailed consideration in the EIS. Moreover, the Commission must explain the basis for each conclusion that further consideration of a suggested alternative is unwarranted." Id., at 331, 547 F. 2d, at 628, quoting from Indiana & Michigan Electric Co. v. FPC, 163 U. S. App. D. C. 334, 337, 502 F. 2d 336, 339 (1974), cert. denied, 420 U. S. 946 (1975).61
While the court's rationale is not entirely unappealing as an abstract proposition, as applied to this case we think it basically misconceives not only the scope of the agency's statutory responsibility, but also the nature of the administrative process, the thrust of the agency's decision, and the type of issues the intervenors were trying to raise.62
There is little doubt that under the Atomic Energy Act of 1954, state public utility commissions or similar bodies are empowered to make the initial decision regarding the need for power. 42 U. S. C. § 2021 (k). The Commission's prime area of concern in the licensing context, on the other hand, is national security, public health, and safety. §§ 2132, 2133, 2201. And it is clear that the need, as that term is conventionally used, for the power was thoroughly explored in the hearings. Even the Federal Power Commission, which regulates  sales in interstate commerce, 16 U. S. C. § 824 et seq. (1976 ed.), agreed with Consumers Power's analysis of projected need. App. 207.63
NEPA, of course, has altered slightly the statutory balance, requiring "a detailed statement by the responsible official on . . . alternatives to the proposed action." 42 U. S. C. § 4332 (C). But, as should be obvious even upon a moment's reflection, the term "alternatives" is not self-defining. To make an impact statement something more than an exercise in frivolous boilerplate the concept of alternatives must be bounded by some notion of feasibility. As the Court of Appeals for the District of Columbia Circuit has itself recognized:64
"There is reason for concluding that NEPA was not meant to require detailed discussion of the environmental effects of `alternatives' put forward in comments when these effects cannot be readily ascertained and the alternatives are deemed only remote and speculative possibilities, in view of basic changes required in statutes and policies of other agencies—making them available, if at all, only after protracted debate and litigation not meaningfully compatible with the time-frame of the needs to which the underlying proposal is addressed." Natural Resources Defense Council v. Morton, 148 U. S. App. D. C. 5, 15-16, 458 F. 2d 827, 837-838 (1972).
See also Life of the Land v. Brinegar, 485 F. 2d 460 (CA9 1973), cert. denied, 416 U. S. 961 (1974). Common sense also teaches us that the "detailed statement of alternatives" cannot be found wanting simply because the agency failed to include every alternative device and thought conceivable by the mind of man. Time and resources are simply too limited to hold that an impact statement fails because the agency failed to ferret out every possible alternative, regardless of how uncommon or unknown that alternative may have been at the time the project was approved.66
 With these principles in mind we now turn to the notion of "energy conservation," an alternative the omission of which was thought by the Court of Appeals to have been "forcefully pointed out by Saginaw in its comments on the draft EIS." 178 U. S. App. D. C., at 328, 547 F. 2d, at 625. Again, as the Commission pointed out, "the phrase `energy conservation' has a deceptively simple ring in this context. Taken literally, the phrase suggests a virtually limitless range of possible actions and developments that might, in one way or another, ultimately reduce projected demands for electricity from a particular proposed plant." App. 331. Moreover, as a practical matter, it is hard to dispute the observation that it is largely the events of recent years that have emphasized not only the need but also a large variety of alternatives for energy conservation. Prior to the drastic oil shortages incurred by the United States in 1973, there was little serious thought in most Government circles of energy conservation alternatives. Indeed, the Council on Environmental Quality did not promulgate regulations which even remotely suggested the need to consider energy conservation in impact statements until August 1, 1973. See 40 CFR § 1500.8 (a) (4) (1977); 38 Fed. Reg. 20554 (1973). And even then the guidelines were not made applicable to draft and final statements filed with the Council before January 28, 1974. Id., at 20557, 21265. The Federal Power Commission likewise did not require consideration of energy conservation in applications to build hydroelectric facilities until June 19, 1973. 18 CFR pt. 2, App. A., § 8.2 (1977); 38 Fed. Reg. 15946, 15949 (1973). And these regulations were not made retroactive either. Id., at 15946. All this occurred over a year and a half after the draft environmental statement for Midland had been prepared, and over a year after the final environmental statement had been prepared and the hearings completed.67
We think these facts amply demonstrate that the concept of "alternatives" is an evolving one, requiring the agency to  explore more or fewer alternatives as they become better known and understood. This was well understood by the Commission, which, unlike the Court of Appeals, recognized that the Licensing Board's decision had to be judged by the information then available to it. And judged in that light we have little doubt the Board's actions were well within the proper bounds of its statutory authority. Not only did the record before the agency give every indication that the project was actually needed, but also there was nothing before the Board to indicate to the contrary.68
We also think the court's criticism of the Commission's "threshold test" displays a lack of understanding of the historical setting within which the agency action took place and of the nature of the test itself. In the first place, while it is true that NEPA places upon an agency the obligation to consider every significant aspect of the environmental impact of a proposed action, it is still incumbent upon intervenors who wish to participate to structure their participation so that it is meaningful, so that it alerts the agency to the intervenors' position and contentions. This is especially true when the intervenors are requesting the agency to embark upon an exploration of uncharted territory, as was the question of energy conservation in the late 1960's and early 1970's.69
"[C]omments must be significant enough to step over a threshold requirement of materiality before any lack of agency response or consideration becomes of concern. The comment cannot merely state that a particular mistake was made . . . ; it must show why the mistake was of possible significance in the results . . . ." Portland Cement Assn. v. Ruckelshaus, 158 U. S. App. D. C. 308, 327, 486 F. 2d 375, 394 (1973), cert. denied sub nom. Portland Cement Corp. v. Administrator, EPA, 417 U. S. 921 (1974).
Indeed, administrative proceedings should not be a game or a forum to engage in unjustified obstructionism by making  cryptic and obscure reference to matters that "ought to be" considered and then, after failing to do more to bring the matter to the agency's attention, seeking to have that agency determination vacated on the ground that the agency failed to consider matters "forcefully presented." In fact, here the agency continually invited further clarification of Saginaw's contentions. Even without such clarification it indicated a willingness to receive evidence on the matters. But not only did Saginaw decline to further focus its contentions, it virtually declined to participate, indicating that it had "no conventional findings of fact to set forth" and that it had not "chosen to search the record and respond to this proceeding by submitting citations of matter which we believe were proved or disproved."71
We also think the court seriously mischaracterized the Commission's "threshold test" as placing "heavy substantive burdens. . . on intervenors . . . ." 178 U. S. App. D. C., at 330, and n. 11, 547 F. 2d, at 627, and n. 11. On the contrary, the Commission explicitly stated:72
"We do not equate this burden with the civil litigation concept of a prima facie case, an unduly heavy burden in this setting. But the showing should be sufficient to require reasonable minds to inquire further." App. 344 n. 27.
We think this sort of agency procedure well within the agency's discretion.74
In sum, to characterize the actions of the Commission as "arbitrary or capricious" in light of the facts then available to it as described at length above, is to deprive those words of any meaning. As we have said in the past:75
"Administrative consideration of evidence . . . always creates a gap between the time the record is closed and the time the administrative decision is promulgated [and, we might add, the time the decision is judicially reviewed]. . . . If upon the coming down of the order  litigants might demand rehearings as a matter of law because some new circumstance has arisen, some new trend has been observed, or some new fact discovered, there would be little hope that the administrative process could ever be consummated in an order that would not be subject to reopening." ICC v. Jersey City, 322 U. S. 503, 514 (1944).
See also Northern Lines Merger Cases, 396 U. S. 491, 521 (1970).77
We have also made it clear that the role of a court in reviewing the sufficiency of an agency's consideration of environmental factors is a limited one, limited both by the time at which the decision was made and by the statute mandating review.78
"Neither the statute nor its legislative history contemplates that a court should substitute its judgment for that of the agency as to the environmental consequences of its actions." Kleppe v. Sierra Club, 427 U. S., at 410 n. 21.
We think the Court of Appeals has forgotten that injunction here and accordingly its judgment in this respect must also be reversed.80
Finally, we turn to the Court of Appeals' holding that the Licensing Board should have returned the ACRS report to ACRS for further elaboration, understandable to a layman, of the reference to other problems.82
The Court of Appeals reasoned that since one function of the report was "that all concerned may be apprised of the safety or possible hazard of the facilities," the report must be in terms understandable to a layman and replete with cross-references to previous reports in which the "other problems" are detailed. Not only that, but if the report does not so elaborate, and the Licensing Board fails to sua sponte return the report to ACRS for further development, the entire agency action, made after exhaustive studies, reviews, and 14 days of hearings, must be nullified.83
Again the Court of Appeals has unjustifiably intruded into the administrative process. It is true that Congress thought publication of the ACRS report served an important function. But the legislative history shows that the function of publication was subsidiary to its main function, that of providing technical advice from a body of experts uniquely qualified to provide assistance. See 42 U. S. C. § 2039; S. Rep. No. 296, 85th Cong., 1st Sess., 24 (1957); Joint Committee on Atomic Energy, A Study of AEC Procedures and Organization in the Licensing of Reactor Facilities, 85th Cong., 1st Sess., 32-34 (Comm. Print 1957). The basic information to be conveyed to the public is not necessarily a full technical exposition of every facet of nuclear energy, but rather the ACRS's position, and reasons therefore, with respect to the safety of a proposed nuclear reactor. Accordingly, the ACRS cannot be faulted for not dealing with every facet of nuclear energy in every report it issues.84
Of equal significance is the fact that the ACRS was not obfuscating its findings. The reports to which it referred were matters of public record, on file in the Commission's  public-documents room. Indeed, all ACRS reports are on file there. Furthermore, we are informed that shortly after the Licensing Board's initial decision, ACRS prepared a list which identified its "generic safety concerns." In light of all this it is simply inconceivable that a reviewing court should find it necessary or permissible to order the Board to sua sponte return the report to ACRS. Our view is confirmed by the fact that the putative reason for the remand was that the public did not understand the report, and yet not one member of the supposedly uncomprehending public even asked that the report be remanded. This surely is, as petitioner Consumers Power claims, "judicial intervention run riot." Brief for Petitioner in No. 76-528, p. 37.85
We also think it worth noting that we find absolutely nothing in the relevant statutes to justify what the court did here. The Commission very well might be able to remand a report for further clarification, but there is nothing to support a court's ordering the Commission to take that step or to support a court's requiring the ACRS to give a short explanation, understandable to a layman, of each generic safety concern.86
All this leads us to make one further observation of some relevance to this case. To say that the Court of Appeals' final reason for remanding is insubstantial at best is a gross understatement. Consumers Power first applied in 1969 for a construction permit—not even an operating license, just a construction permit. The proposed plant underwent an incredibly extensive review. The reports filed and reviewed literally fill books. The proceedings took years, and the actual hearings themselves over two weeks. To then nullify that effort seven years later because one report refers to other problems, which problems admittedly have been discussed at length in other reports available to the public, borders on the Kafkaesque. Nuclear energy may some day be a cheap, safe source of power or it may not. But Congress has made a  choice to at least try nuclear energy, establishing a reasonable review process in which courts are to play only a limited role. The fundamental policy questions appropriately resolved in Congress and in the state legislatures are not subject to re-examination in the federal courts under the guise of judicial review of agency action. Time may prove wrong the decision to develop nuclear energy, but it is Congress or the States within their appropriate agencies which must eventually make that judgment. In the meantime courts should perform their appointed function. NEPA does set forth significant substantive goals for the Nation, but its mandate to the agencies is essentially procedural. See 42 U. S. C. § 4332. See also Aberdeen & Rockfish R. Co. v. SCRAP, 422 U. S., at 319. It is to insure a fully informed and well-considered decision, not necessarily a decision the judges of the Court of Appeals or of this Court would have reached had they been members of the decisionmaking unit of the agency. Administrative decisions should be set aside in this context, as in every other, only for substantial procedural or substantive reasons as mandated by statute, Consolo v. FMC, 383 U. S. 607, 620 (1966), not simply because the court is unhappy with the result reached. And a single alleged oversight on a peripheral issue, urged by parties who never fully cooperated or indeed raised the issue below, must not be made the basis for overturning a decision properly made after an otherwise exhaustive proceeding.87
Reversed and remanded.88
MR. JUSTICE BLACKMUN and MR. JUSTICE POWELL took no part in the consideration or decision of these cases.89
 Together with No. 76-528, Consumers Power Co. v. Aeschliman et al., also on certiorari to the same court.90
 Briefs of amici curiae urging reversal were filed by Cameron F. MacRae, Leonard M. Trosten, and Harry H. Voigt for Edison Electric Institute et al. in No. 76-419; by Leonard J. Theberge, John M. Cannon, Edward H. Dowd, and L. Manning Muntzing for Hans A. Bethe et al. in No. 76-528; and by Max Dean and David S. Heller for the U. S. Labor Party in No. 76-528.91
Louis J. Lefkowitz, Attorney General of New York, Samuel A. Hirshowitz, First Assistant Attorney General, Philip Weinberg and John F. Shea III, Assistant Attorneys General; Cabanne Howard, Assistant Attorney General of Maine; and Ellyn Weiss, Assistant Attorney General of Massachusetts, field a brief for 24 named States as amici curiae urging affirmance in both cases, joined by officials for their respective States as follows: William J. Baxley, Attorney General of Alabama, and Henry H. Caddell, Assistant Attorney General; Richard R. Wier, Jr., Attorney General of Delaware, and June D. MacArtor, Deputy Attorney General; Robert L. Shevin, Attorney General of Florida, and Marty Friedman, Assistant Attorney General; Arthur K. Bolton, Attorney General of Georgia, and Robert Bomar, Senior Assistant Attorney General; William J. Scott, Attorney General of Illinois, and Richard W. Cosby, Assistant Attorney General; Curt T. Schneider, Attorney General of Kansas, and William Griffin, Assistant Attorney General; Robert F. Stephens, Attorney General of Kentucky, and David Short, Assistant Attorney General; William J. Guste, Attorney General of Louisiana, and Richard M. Troy, Assistant Attorney General; Joseph E. Brennan, Attorney General of Maine; Francis B. Burch, Attorney General of Maryland, and Warren K. Rich, Assistant Attorney General; Francis X. Bellotti, Attorney General of Massachusetts; Frank J. Kelley, Attorney General of Michigan, and Stewart H. Freeman, Assistant Attorney General; Warren R. Spannaus, Attorney General of Minnesota, and Jocelyn F. Olson, Assistant Attorney General; John Ashcroft, Attorney General of Missouri, and Robert H. Lindholm, Assistant Attorney General; Toney Anaya, Attorney General of New Mexico, and James Huber, Assistant Attorney General; Rufus L. Edmisten, Attorney General of North Carolina, and Dan Oakley, Assistant Attorney General; William J. Brown, Attorney General of Ohio, and David Northrup, Assistant Attorney General; James A. Redden, Attorney General of Oregon, and Richard M. Sandvik, Assistant Attorney General; Robert P. Kane, Attorney General of Pennsylvania, and Douglas Blazey, Assistant Attorney General; John L. Hill, Attorney General of Texas, and Troy C. Webb and Paul G. Gosselink, Assistant Attorneys General; Robert B. Hansen, Attorney General of Utah, and William C. Quigley; M. Jerome Diamond, Attorney General of Vermont, and Benson D. Scotch, Assistant Attorney General; and Bronson C. LaFollette, Attorney General of Wisconsisn, and John E. Kofron, Assistant Attorney General. George C. Deptula and James N. Barnes filed a brief for the Union of Concerned Scientists Fund, Inc., as amicus curiae urging affirmance in No. 76-419.92
Ronald A. Zumbrun, Raymond M. Momboisse, Robert K. Best, Albert Ferri, Jr., and W. Hugh O'Riordan filed a brief for the Pacific Legal Foundation as amicus curiae in both cases.93
 While there was division in this Court in United States v. Florida East Coast R. Co. with respect to the constitutionality of such an interpretation in a case involving ratemaking, which Mr. Justice Douglas and MR. JUSTICE STEWART felt was "adjudicatory" within the terms of the Act, the cases in the Court of Appeals for the District of Columbia Circuit which we review here involve rulemaking procedures in their most pristine sense.94
 The licensing and regulatory functions of the Atomic Energy Commission (AEC) were transferred to the Nuclear Regulatory Commission (NRC) by the Energy Reorganization Act of 1974, 42 U. S. C. § 5801 et seq. (1970 ed., Supp. V). Hereinafter both the AEC and NRC will be referred to as the Commission.95
 ACRS is required to review each construction permit application for the purpose of informing the Commission of the "hazards of proposed or existing reactor facilities and the adequacy of proposed reactor safety standards." 42 U. S. C. § 2039.96
 The Licensing Board issues a permit if it concludes that there is reasonable assurance that the proposed plant can be constructed and operated without undue risk, 42 U. S. C. § 2241; 10 CFR § 50.35 (a) (1977), and that the environmental cost-benefit balance favors the issuance of a permit.97
 When a license application is contested, the Licensing Board must find reasonable assurance that the plant can be operated without undue risk and will not be inimical to the common defense and security or to the health and safety of the public. See 42 U. S. C. § 2232 (a); 10 CFR § 50.57 (a) (1977). The Licensing Board's decision is subject to review similar to that afforded the Board's decision with respect to a construction permit.98
 The nuclear fission which takes place in light-water nuclear reactors apparently converts its principal fuel, uranium, into plutonium, which is itself highly radioactive but can be used as reactor fuel if separated from the remaining uranium and radioactive waste products. Fuel reprocessing refers to the process necessary to recapture usable plutonium. Waste disposal, at the present stage of technological development, refers to the storage of the very long lived and highly radioactive waste products until they detoxify sufficiently that they no longer present an environmental hazard. There are presently no physical or chemical steps which render this waste less toxic, other than simply the passage of time.99
 The Commission stated:100
"In our view, the procedures adopted provide a more than adequate basis for formulation of the rule we adopted. All parties were fully heard. Nothing offered was excluded. The record does not indicate that any evidentiary material would have been received under different procedures. Nor did the proponent of the strict `adjudicatory' approach make an offer of proof—or even remotely suggest—what substantive matters it would develop under different procedures. In addition, we note that 11 documents including the Survey were available to the parties several weeks before the hearing, and the Regulatory staff, though not requested to do so, made available various drafts and handwritten notes. Under all of the circumstances, we conclude that adjudicatory type procedures were not warranted here." App. 389-390 (footnote omitted).
 The ACRS report as quoted, 178 U. S. App. D. C., at 333, 547 F. 2d, at 630, stated:102
"Other problems related to large water reactors have been identified by the Regulatory Staff and the ACRS and cited in previous ACRS reports. The Committee believes that resolution of these items should apply equally to the Midland Plant Units 1 & 2.
"The Committee believes that the above items can be resolved during construction and that, if due consideration is given to these items, the nuclear units proposed for the Midland Plant can be constructed with reasonable assurance that they can be operated without undue risk to the health and safety of the public."
 Saginaw included the Saginaw Valley Nuclear Study Group, the Citizens Committee for Environmental Protection of Michigan, the United Automobile Workers International, and three other environmental groups. Mapleton included Nelson Aeschliman and five other residents of a community near the proposed plantsite. Mapleton did not raise any contentions relating to energy conservation.104
 Pursuant to the regulations then in effect, the Licensing Board refused to consider most of the environmental issues in this first set of hearings. On the last day of those hearings, however, the Court of Appeals for the District of Columbia Circuit decided Calvert Cliffs' Coordinating Comm. v. AEC, 146 U. S. App. D. C. 33, 449 F. 2d 1109 (1971), which invalidated the Commission's NEPA regulations. One effect of that decision was to require that environmental matters be considered in pending proceedings, including this one. Accordingly, the Commission revised its regulations and then undertook an extensive environmental review of the proposed nuclear plants, requiring Consumers Power to file a lengthy environmental report. Thereafter the Commission's staff prepared the draft environmental impact statement discussed in text.105
 The Licensing Board had highlighted this same problem in its initial decision, noting "that the failure to propose proper findings and conclusions has greatly complicated the task of the Board and has made it virtually impossible in some instances to know whether particular issues are in fact contested." App. 190 n. 10. The Appeal Board was even less charitable, noting that that "[p]articipation in this manner, in our opinion, subverts the entire adjudicatory process." Id., at 257.106
 In what was essentially dictum, the Commission also ruled, after considering the various relevant factors—such as the extent to which the new rule represents a departure from prior practice, the degree of reliance on past practice and consequent burdens imposed by retroactive application of the rule—that the rule enunciated in Niagara should not be applied retroactively to cases which had progressed to final order and issuance of construction permits before Niagara was decided. App. 337.107
 In the Court of Appeals no one questioned the Commission's authority to deal with fuel cycle issues by informal rulemaking as opposed to adjudication. 178 U. S. App. D. C., at 345-346, 547 F. 2d, at 642-643. Neither does anyone seriously question before this Court the Commission's authority in this respect.108
 After the decision of the Court of Appeals the Commission promulgated a new interim rule pending issuance of a final rule. 42 Fed. Reg. 13803 (1977). See Vermont Yankee Nuclear Power Corp., 5 N. R. C. 717 (1977). The Commission then, at the request of the New England Coalition on Nuclear Pollution, applied the interim rule to Vermont Yankee and determined that the cost-benefit analysis was still in the plant's favor. Vermont Yankee Nuclear Power Corp., 6 N. R. C. 25 (1977). That decision is presently on appeal to the Court of Appeals for the First Circuit. The Commission has also indicated in its brief that it intends to complete the proceedings currently in progress looking toward the adoption of a final rule regardless of the outcome of this case. Brief for Federal Respondents 37 n. 36. Following oral argument, respondent NRDC, relying on the above facts, filed a suggestion of mootness and a motion to dismiss the writ of certiorari as improvidently granted. We hold that the case is not moot, and deny the motion to dismiss the writ of certiorari as improvidently granted.109
Upon remand, the majority of the panel of the Court of Appeals is entirely free to agree or disagree with Judge Tamm's conclusion that the rule pertaining to the back end of the fuel cycle under which petitioner Vermont Yankee's license was considered is arbitrary and capricious within the meaning of § 10 (e) of the Administrative Procedure Act, 5 U. S. C. § 706 (1976 ed.), even though it may not hold, as it did in its previous opinion, that the rule is invalid because of the inadequacy of the agency procedures. Should it hold the rule invalid, it appears in all probability that the Commission will proceed to promulgate a rule resulting from rulemaking proceedings currently in progress. Brief for Federal Respondents 37 n. 36. In all likelihood the Commission would then be required, under the compulsion of the court's order, to examine Vermont Yankee's license under that new rule.110
If, on the other hand, a majority of the Court of Appeals should decide that it was unwilling to hold the rule in question arbitrary and capricious merely on the basis of § 10 (e) of the Administrative Procedure Act, Vermont Yankee would not necessarily be required to have its license reevaluated. So far as petitioner Vermont Yankee is concerned, there is certainly a case or controversy in this Court with respect to whether it must, by virtue of the Court of Appeals' decision, submit its license to the Commission for reevaluation and possible revocation under a new rule. It is true that we do not finally determine here the validity of the rule upon which the validity of Vermont Yankee's license in turn depends. Neither should anything we say today be taken as a limitation on the Court of Appeals' discretion to take due account, if appropriate, of any additions made to the record by the Commission or to consolidate this appeal with the appeal from the interim rulemaking proceeding which is already pending. But the fact that the question of the validity of the first rule remains open upon remand makes the controversy no less "live."111
As we read the opinion of the Court of Appeals, its view that reviewing courts may in the absence of special circumstances justifying such a course of action impose additional procedural requirements on agency action raises questions of such significance in this area of the law as to warrant our granting certiorari and deciding the case. Since the vast majority of challenges to administrative agency action are brought to the Court of Appeals for the District of Columbia Circuit, the decision of that court in this case will serve as precedent for many more proceedings for judicial review of agency actions than would the decision of another Court of Appeals. Finally, this decision will continue to play a major role in the instant litigation regardless of the Commission's decision to press ahead with further rulemaking proceedings. As we note in n. 15, infra, not only is the NRDC relying on the decision of the Court of Appeals as a device to force the agency to provide more procedures, but it is also challenging the interim rules promulgated by the agency in the Court of Appeals, alleging again the inadequacy of the procedures and citing the opinion of the Court of Appeals as binding precedent to that effect.112
 Vermont Yankee's interpretation has been consistent throughout the litigation. That cannot be said of the other parties, however. The Government, Janus-like, initially took both positions. While the petition for certiorari was pending, a brief was filed on behalf of the United States and the Commission, with the former indicating that it believed the court had unanimously held the record to be inadequate, while the latter took Vermont Yankee's view of the matter. See Brief for Federal Respondents 5-9 (filed Jan. 10, 1977). When announcing its intention to undertake licensing of reactors pending the promulgation of an "interim" fuel cycle rule, however, the Commission said:113
"[T]he court found that the rule was inadequately supported by the record insofar as it treated two particular aspects of the fuel cycle—the impacts from reprocessing of spent fuel and the impacts from radioactive waste management." 41 Fed. Reg. 45850 (1976).
And even more recently, in opening another rulemaking proceeding to replace the rule overturned by the Court of Appeals, the Commission stated:115
"The original procedures proved adequate for the development and illumination of a wide range of fuel cycle impact issues . . . .
". . . The court here indicated that the procedures previously employed could suffice, and indeed did for other issues.
"Accordingly, notice is hereby given that the rules for the conduct of the reopened hearing and the authorities and responsibilities of the Hearing Board will be the same as originally applied in this matter (38 Fed. Reg. 49, January 3, 1973) except that specific provision is hereby made for the Hearing Board to entertain suggestions from participants as to questions which the Board should ask of witnesses for other participants." 42 Fed. Reg. 26988-26989 (1977).
Respondent NRDC likewise happily switches sides depending on the forum. As indicated above, it argues here that the Court of Appeals held only that the record was inadequate. Almost immediately after the Court of Appeals rendered its decision, however, NRDC filed a petition for rulemaking with the Commission which listed over 13 pages of procedural suggestions it thought "necessary to comply with the Court's order and with the mandate of [NEPA]." NRDC, Petition for Rulemaking, NRC Docket No. RM-50-3 (Aug. 10, 1976). These proposals include cross-examination, discovery, and subpoena power. Id., Attachment, Rules for Conduct of Hearing on Environmental Effects of the Uranium Fuel Cycle, ¶¶ 5 (a), 9 (b), 11. NRDC likewise challenged the interim fuel cycle rule and suggested to the Court of Appeals that it hold the case pending our decision in this case because the interim rules were "defective due to the inadequacy of the procedures used in developing the rule . . . ." Motion to Hold Petition for Review in Abeyance 1, in NRDC v. NRC, No. 77-1448 (DC Cir., petition for review filed May 13, 1977; motion filed July 5, 1977). NRDC has likewise challenged the procedures being used in the final rulemaking proceeding as being "no more than a re-run of hearing procedures which were found inadequate [by the Court of Appeals]." NRDC Petition for Reconsideration of the Ruling Reopening the Hearings on the Environmental Effects of the Uranium Fuel Cycle 10, NRC Docket No. RM-50-3 (June 6, 1977).117
 Respondent NRDC does not now argue that additional procedural devices were required under the Constitution. Since this was clearly a rulemaking proceeding in its purest form, we see nothing to support such a view. See United States v. Florida East Coast R. Co., 410 U. S. 224, 244-245 (1973); Bowles v. Willingham, 321 U. S. 503 (1944); Bi-Metallic Investment Co. v. State Board of Equalization, 239 U. S. 441 (1915).118
 NRDC argues that the agency has in the past provided more than the minimum procedures specified in § 4 of the APA and therefore something more is required here, since "[a]gencies are not free to alter their procedures on a whim, grossly constricting parties' procedural rights when it deems them an impediment or embarrassment to implementing its own views." Brief for Respondents in No. 76-419, p. 46. In support NRDC first argues that the Commission has considered other equally generic issues in adjudicatory proceedings. But NRDC conceded in the court below that the agency could promulgate rules regarding the fuel cycle in rulemaking proceedings. 178 U. S. App. D. C., at 346, 547 F. 2d, at 643. Moreover, even here it concedes "that the Commission has in the past chosen to consider both environmental and safety issues that would ordinarily be addressed in adjudicatory licensing proceedings through `generic' rulemaking, a practice with which the lower court did not take issue." Brief for Respondents in No. 76-419, p. 48. It now contends, however, that the Commission provided more procedural safeguards in those rulemaking proceedings than in the proceeding presently under review. In support it cites three previous proceedings where cross-examination was supposedly provided. Id., at 49 n. 69.119
Pretermitting both the fact that the Court of Appeals in no way relied upon this argument in its decision and the question of whether courts can impose additional procedures even when an agency substantially departs from past practice, we find NRDC's argument without merit. In the first place, three proceedings out of the many held by NRC and its predecessor hardly establish the type of longstanding and well-established practice deviation from which might justify judicial intervention. It appears, moreover, that in fact the hearings cited by NRDC are not only not part of a longstanding practice but are themselves aberrational. Since 1970 the Commission has conducted a large number of rulemaking proceedings, some of which have involved matters of substantial importance, and almost none of which have involved cross-examination. See, e. g., Quality Assurance Criteria for Nuclear Power Plants, 35 Fed. Reg. 10499 (1970); General Design Criteria for Nuclear Power Plants, 36 Fed. Reg. 3255 (1971); Pre-Construction Permit Activities, 39 Fed. Reg. 14506 (1974); Environmental Protection—Licensing and Regulatory Policy and Procedures. Id., at 26279.120
 See, e. g., CAB v. Hermann, 353 U. S. 322 (1957); Oklahoma Press Pub. Co. v. Walling, 327 U. S. 186 (1946); Wallace Corp. v. NLRB, 323 U. S. 248 (1944); Endicott Johnson Corp. v. Perkins, 317 U. S. 501 (1943); Utah Fuel Co. v. National Bituminous Coal Comm'n, 306 U. S. 56 (1939); Norwegian Nitrogen Products Co. v. United States, 288 U. S. 294 (1933).121
 See Power Reactor Co. v. Electricians, 367 U. S. 396, 408 (1961); United States v. Zucca, 351 U. S. 91, 96 (1956).122
 See Wright, The Courts and the Rulemaking Process: The Limits of Judicial Review, 59 Cornell L. Rev. 375, 387-388 (1974).123
 Of course, the court must determine whether the agency complied with the procedures mandated by the relevant statutes. Citizens to Preserve Overton Park v. Volpe, 401 U. S. 402, 417 (1971). But, as we indicated above, there is little doubt that the agency was in full compliance with all the applicable requirements of the Administrative Procedure Act.124
 The court also indicated at the end of the opinion in Aeschliman that since "this matter requires remand and reopening of the issues of energy conservation alternatives as well as recalculation of costs and benefits, we assume that the Commission will take into account the changed circumstances regarding Dow's [the principal customer for the plant's steam] need for process steam, and the intended continued operation of Dow's fossil-fuel generating facilities." 178 U. S. App. D. C., at 335, 547 F. 2d, at 632. As we read the Court of Appeals opinion, however, this was not an independent basis for vacating and remanding the Commission's licensing decision. It also appears from the record that the Commission has reconsidered the changed circumstances and refused to reopen the proceedings at least three times, see App. 346-347, 348-349, 350-351, and possibly a fourth, see Brief for Nonfederal Respondents in No. 76-528, pp. 19-20, n. 8. We see no error in the Commission's actions in this respect.
568 F.2d 240 (1977)2
No. 758, Docket 76-6169.
United States Court of Appeals, Second Circuit.
Argued May 5, 1977.
Decided December 15, 1977.
  Joseph H. Einstein, New York City (Aranow, Brodsky, Bohlinger, Benetar & Einhorn, New York City, of counsel), for defendants-appellants.5
David W. McMorrow, Asst. U. S. Atty., E.D.N.Y., Brooklyn, N. Y., (David G. Trager, U. S. Atty., Alvin A. Schall, Asst. U. S. Atty., E.D.N.Y., Brooklyn, N. Y., and Arnold I. Friede, Asst. Chief Counsel, U. S. Food & Drug Admin., Washington, D. C., of counsel), for plaintiff-appellee.6
Richard S. Morey, Washington, D. C. (Kleinfeld, Kaplan & Becker, Washington, D. C., of counsel), for intervenor-appellant.7
Before WATERMAN and GURFEIN, Circuit Judges, and BLUMENFELD, District Judge.8
This appeal involving a regulation of the Food and Drug Administration is not here upon a direct review of agency action. It is an appeal from a judgment of the District Court for the Eastern District of New York (Hon. John J. Dooling, Judge) enjoining the appellants, after a hearing, from processing  hot smoked whitefish except in accordance with time-temperature-salinity (T-T-S) regulations contained in 21 C.F.R. Part 122 (1977). The thorough analytical opinion of the District Court is reported at 417 F.Supp. 1364 (Aug. 17, 1976).10
The injunction was sought and granted on the ground that smoked whitefish which has been processed in violation of the T-T-S regulation is "adulterated." Food, Drug and Cosmetics Act ("the Act") §§ 302(a) and 301(k), 21 U.S.C. §§ 332(a), 331(k).11
Appellant Nova Scotia receives frozen or iced whitefish in interstate commerce which it processes by brining, smoking and cooking. The fish are then sold as smoked whitefish.12
The regulations cited above require that hot-process smoked fish be heated by a controlled heat process that provides a monitoring system positioned in as many strategic locations in the oven as necessary to assure a continuous temperature through each fish of not less than 180° F. for a minimum of 30 minutes for fish which have been brined to contain 3.5% water phase salt or at 150° F. for a minimum of 30 minutes if the salinity was at 5% water phase. Since each fish must meet these requirements, it is necessary to heat an entire batch of fish to even higher temperatures so that the lowest temperature for any fish will meet the minimum requirements.13
Government inspection of appellants' plant established without question that the minimum T-T-S requirements were not being met. There is no substantial claim that the plant was processing whitefish under "insanitary conditions" in any other material respect. Appellants, on their part, do not defend on the ground that they were in compliance, but rather that the requirements could not be met if a marketable whitefish was to be produced. They defend upon the grounds that the regulation is invalid (1) because it is beyond the authority delegated by the statute; (2) because the FDA improperly relied upon undisclosed evidence in promulgating the regulation and because it is not supported by the administrative record; and (3) because there was no adequate statement setting forth the basis of the regulation. We reject the contention that the regulation is beyond the authority delegated by the statute, but we find serious inadequacies in the procedure followed in the promulgation of the regulation and hold it to be invalid as applied to the appellants herein.14
The hazard which the FDA sought to minimize was the outgrowth and toxin formation of Clostridium botulinum Type E spores of the bacteria which sometimes inhabit fish. There had been an occurrence of several cases of botulism traced to consumption of fish from inland waters in 1960 and 1963 which stimulated considerable bacteriological research. These bacteria can be present in the soil and water of various regions. They can invade fish in their natural habitat and can be further disseminated in the course of evisceration and preparation of the fish for cooking. A failure to destroy such spores through an adequate brining, thermal, and refrigeration process was found to be dangerous to public health.15
 The Commissioner of Food and Drugs ("Commissioner"), employing informal "notice-and-comment" procedures under 21 U.S.C. § 371(a), issued a proposal for the control of C. botulinum bacteria Type E in fish. 34 F.R. 17,176 (Oct. 23, 1969). For his statutory authority to promulgate the regulations, the Commissioner specifically relied only upon § 342(a)(4) of the Act which provides:16
"A food shall be deemed to be adulterated —
"(4) if it has been prepared, packed, or held under insanitary conditions whereby it may have become contaminated with filth, or whereby it may have been rendered injurious to health;"
Similar guidelines for smoking fish had been suggested by the FDA several years earlier, and were generally made known to people in the industry. At that stage, however, they were merely guidelines without substantive effect as law. Responding to the Commissioner's invitation in the notice of proposed rulemaking, members of the industry, including appellants and the intervenor-appellant, submitted comments on the proposed regulation.18
The Commissioner thereafter issued the final regulations in which he adopted certain suggestions made in the comments, including a suggestion by the National Fisheries Institute, Inc. ("the Institute"), the intervenor herein. 35 F.R. 17,401 (Nov. 13, 1970). The original proposal provided that the fish would have to be cooked to a temperature of 180° F. for at least 30 minutes, if the fish have been brined to contain 3.5% water phase salt, with no alternative. In the final regulation, an alternative suggested by the intervenor "that the parameter of 150° F. for 30 minutes and 5% salt in the water phase be established as an alternate procedure to that stated in the proposed regulation for an interim period until specific parameters can be established" was accepted, but as a permanent part of the regulation rather than for an interim period.19
The intervenor suggested that "specific parameters" be established. This referred to particular processing parameters for different species of fish on a "species by species" basis. Such "species by species" determination was proposed not only by the intervenor but also by the Bureau of Commercial Fisheries of the Department of the Interior. That Bureau objected to the general application of the T-T-S requirement proposed by the FDA on the ground that application of the regulation to all species of fish being smoked was not commercially feasible, and that the regulation should therefore specify time-temperature-salinity requirements, as developed by research and study, on a species-by-species basis. The Bureau suggested that "wholesomeness considerations could be more practically and adequately realized by reducing processing temperature and using suitable concentrations of nitrite and salt." The Commissioner took cognizance of the suggestion, but decided, nevertheless, to impose the T-T-S requirement on all species of fish (except chub, which were regulated by 21 C.F.R. 172.177 (1977) [dealing with food additives]).20
He did acknowledge, however, in his "basis and purpose" statement required by the Administrative Procedure Act ("APA"), 5 U.S.C. § 553(c), that "adequate times, temperatures and salt concentrations have not been demonstrated for each individual species of fish presently smoked". 35 F.R. 17,401 (Nov. 13, 1970). The Commissioner concluded, nevertheless, that "the processing requirements of the proposed regulations are the safest now known to prevent  the outgrowth and toxin formation of C. botulinum Type E". He determined that "the conditions of current good manufacturing practice for this industry should be established without further delay." Id.21
The Commissioner did not answer the suggestion by the Bureau of Fisheries that nitrite and salt as additives could safely lower the high temperature otherwise required, a solution which the FDA had accepted in the case of chub. Nor did the Commissioner respond to the claim of Nova Scotia through its trade association, the Association of Smoked Fish Processors, Inc., Technical Center that "[t]he proposed process requirements suggested by the FDA for hot processed smoked fish are neither commercially feasible nor based on sound scientific evidence obtained with the variety of smoked fish products to be included under this regulation." (Exhibit D, Tab A).22
Nova Scotia, in its own comment, wrote to the Commissioner that "the heating of certain types of fish to high temperatures will completely destroy the product". It suggested, as an alternative, that "specific processing procedures could be established for each species after adequate work and experimention [sic] has been done — but not before." (Id.). We have noted above that the response given by the Commissioner was in general terms. He did not specifically aver that the T-T-S requirements as applied to whitefish were, in fact, commercially feasible.23
When, after several inspections and warnings, Nova Scotia failed to comply with the regulation, an action by the United States Attorney for injunctive relief was filed on April 7, 1976, six years later, and resulted in the judgment here on appeal. The District Court denied a stay pending appeal, and no application for a stay was made to this court.24
The argument that the regulation is not supported by statutory authority cannot be dismissed out of hand. See Schilling v. Rogers, 363 U.S. 666, 676-77, 80 S.Ct. 1288, 1295-96, 4 L.Ed.2d 1478 (1960); see Batterton v. Francis, 432 U.S. 416, 97 S.Ct. 2399, 53 L.Ed.2d 448 (1977). The sole statutory authority relied upon is § 342(a)(4), quoted above. As we were instructed in S. E. C. v. Chenery Corp., 318 U.S. 80, 87, 63 S.Ct. 454, 459, 87 L.Ed. 626 (1943) (Chenery I), "[t]he grounds upon which an administrative order must be judged are those upon which the record discloses that its action was based." Nor is the Commissioner's expressed reliance solely on § 342(a)(4) a technicality which might be removed by a later and wiser reliance on another subsection. For in this case, as the agency recognized, there is no other section or subsection that can pass as statutory authority for the regulation. The categories of "adulteration" prohibited in Section 342 all refer to food as an "adulterated" product rather than to the process of preparing food, except for subsection (a)(4) which alone deals with the processing of food.26
Appellants contend that the prohibition against "insanitary conditions" embraces conditions only in the plant itself, but does not include conditions which merely inhibit the growth of organisms already in the food when it enters the plant in its raw state. They distinguish between conditions which are insanitary, which they concede to be within the ambit of § 342(a)(4), and conditions of sterilization required to destroy micro-organisms, which they contend are not.27
It is true that on a first reading the language of the subsection appears to cover only "insanitary conditions" "whereby it [the food] may have been rendered injurious to health" (emphasis added). And a plausible argument can, indeed, be made that the references are to insanitary conditions in the plant itself, such as the presence of rodents or insects, e. g., United States v. Park, 421 U.S. 658, 95 S.Ct. 1903, 44 L.Ed.2d 489 (1975); United States v.  Cassaro, Inc., 443 F.2d 153 (1st Cir. 1971); United States v. Hammond Milling Co., 413 F.2d 608 (5th Cir. 1969), cert. denied, 396 U.S. 1002, 90 S.Ct. 552, 24 L.Ed.2d 494 (1970).28
Yet, when we are dealing with the public health, the language of the Food, Drug and Cosmetic Act should not be read too restrictively, but rather as "consistent with the Act's overriding purpose to protect the public health". United States v. Bacto-Unidisk, 394 U.S. 784, 798, 89 S.Ct. 1410, 1418, 22 L.Ed.2d 726 (1969). As Justice Frankfurter said in United States v. Dotterweich, 320 U.S. 277, 280, 64 S.Ct. 134, 136, 88 L.Ed. 48 (1943):29
"The purposes of this legislation thus touch phases of the lives and health of people which, in the circumstances of modern industrialism, are largely beyond self-protection. Regard for these purposes should infuse construction of the legislation if it is to be treated as a working instrument of government and not merely as a collection of English words."
Thus, a provision concerning "food additives" has been held to include even poisonous substances which have not been "added" by human hands. United States v. Ewig Bros. Co., 502 F.2d 715, 721-24 (7th Cir. 1974) (Stevens, J. now Mr. Justice Stevens), cert. denied, 420 U.S. 945, 95 S.Ct. 1325, 43 L.Ed.2d 423 (1975).31
Section 371(a), applicable to rulemaking under § 342(a)(4), provides: "The authority to promulgate regulations for the efficient enforcement of this chapter, except as otherwise provided in this section, is vested in the Secretary." We read this grant as analogous to the provision "make . . . such rules and regulations as may be necessary to carry out the provisions of this Act," in which case "the validity of a regulation promulgated thereunder will be sustained so long as it is `reasonably related to the purposes of the enabling legislation.' [citations omitted]" Mourning v. Family Publications Service, Inc., 411 U.S. 356, 369, 93 S.Ct. 1652, 36 L.Ed.2d 318 (1973). When agency rulemaking serves the purposes of the statute, courts should refuse to adopt a narrow construction of the enabling legislation which would undercut the agency's authority to promulgate such rules. United States v. Midwest Video Corp., 406 U.S. 649, 92 S.Ct. 1860, 32 L.Ed.2d 390 (1972) (upholding FCC's authority to promulgate "cablecasting" regulation as "reasonably ancillary to the effective performance of the Commission's various responsibilities for the regulation of television broadcasting"); and United States v. Storer Broadcasting Co., 351 U.S. 192, 203-04, 76 S.Ct. 763, 100 L.Ed.2d 1081 (1956). The court's role should be one of constructive cooperation with the agency in furtherance of the public interest. International Harvester Co. v. Ruckelhaus, 155 U.S.App.D.C. 411, 443, 478 F.2d 615, 647 (1973). As the Supreme Court has said, "We are, in the absence of compelling evidence that such was Congress' intention, unwilling to prohibit administrative action imperative for the achievement of an agency's ultimate purposes." In re Permian Basin Area Rate Cases, 390 U.S. 747, 780, 88 S.Ct. 1344, 20 L.Ed.2d 312 (1968).32
Appellant's argument, it should be noted, is not that there has been an unlawful delegation of legislative power, Schechter Poultry Corp. v. United States, 295 U.S. 495, 55 S.Ct. 837, 79 L.Ed. 1570 (1935), or even a delegation of "unfettered discretion." See Amalgamated Meat Cutters & Butcher Workers v. Connally, 337 F.Supp. 737, 757 (D.D.C.1971) (Leventhal, J.). The argument, fairly construed, is that Congress did not mean to go so far as to require sterilization sufficient to kill bacteria that may be in the food itself rather than bacteria which accreted in the factory through the use of insanitary equipment.33
 There are arguments which can indeed be mustered to support such a broad-based attack under 5 U.S.C. § 706.34
First, the Act deals with standards of identity and various categories that can render food harmful to health. Yet, so far as the category of harmful micro-organisms is concerned, there is only a single provision, 21 U.S.C. § 344, which directly deals with "micro-organisms." That provision is limited to emergency permit controls dealing with any class of food which the Secretary finds, after investigation, "may, by reason of contamination with micro-organisms during the manufacture, processing or packing thereof in any locality, be injurious to health, and that such injurious nature cannot be adequately determined after such articles have entered interstate commerce, [in which event] he then, and in such case only, shall promulgate regulations providing for the issuance . . . of permits . . .." (Emphasis added.) It may be argued that the failure to mention "micro-organisms" in the "adulteration" section of the Act, which includes § 342(a)(4), means that Congress intended to delegate no further authority to control micro-organisms than is expressed in the "emergency" control of Section 344.35
On the other hand, as Judge Dooling held, the manner of processing can surely give rise to the survival, with attendant toxic effects on humans, of spores which would not have survived under stricter "sanitary" conditions. In that sense, treating "insanitary conditions" in relation to the hazard, the interpretation of the District Court which described the word "sanitary" as merely "inelegant" is a fair reading, emphasizing that the food does not have to be actually contaminated during processing and packing but simply that "it may have been rendered injurious to health," § 342(a)(4), by inadequate sanitary conditions of prevention.36
The other argument of some force would involve the difference in agency rulemaking procedure which results from treating the regulations in question as supported by § 342(a)(4). The Act was enacted in substantially its present form in 1938, preceding the APA by about eight years. In 1938 Congress, groping for standards for rulemaking and the scope of judicial review (as it still is) created a bifurcated structure for rulemaking and, hence, for judicial review.37
Congress decided to allow informal rulemaking to the FDA generally, § 371(a), but it also provided for formal rulemaking which, upon request, required "a public hearing for the purpose of receiving evidence relevant and material to the issues raised by such objections." § 371(e). It further provided that in such cases the "order shall be based only on substantial evidence of record at such hearing and shall set forth, as part of the order, detailed findings of fact on which the order is based."38
The formal rulemaking with its concomitant standard of "substantial evidence of record," see United States v. Florida East Coast Railway Co., 410 U.S. 224, 93 S.Ct. 810, 35 L.Ed.2d 223 (1973), is limited, however, to rulemaking under specific enumerated sections of the Act. Section 342, the "adulteration" section upon which the Commissioner relied for his statutory authority is not one of these. The section dealing with temporary permits for micro-organisms, § 344, is. See § 371(e)(1). Thus, a temporary suspension because of the presence of micro-organisms in food merits a formal procedure while permanent regulation of micro-organisms is achievable by informal "notice-and-comment" procedure. Even though we read the statute § 342(a)(4) broadly in terms of the authority delegated to the agency, we must, nevertheless, view with some strictness the minimal requirements for the informal "notice and comment" procedure that follows as of course — a matter we shall discuss below.39
We do not discount the logical arguments in support of a restrictive reading of  § 342(a)(4), but we perceive a larger general purpose on the part of Congress in protecting the public health.40
We come to this conclusion, aside from the general rules of construction noted above, for several reasons: First, until this enforcement proceeding was begun, no lawyer at the knowledgeable Food and Drug bar ever raised the question of lack of statutory delegation or even hinted at such a question. Second, the body of data gathered by the experts, including those of the Technical Laboratory of the Bureau of Fisheries manifested a concern about the hazards of botulism. Third, analogously, the Meat Inspection Act of 1907 (now codified as amended at 21 U.S.C. § 608), which hardly provided a clearer standard than does the "insanitary conditions" provision in the Food and Drug Act, has regulations under it concerning mandatory temperatures for processing pork muscle tissue to eliminate the hazard of trichonosis. The statute permits the Secretary "to prescribe the rules and regulations of sanitation under which such establishments shall be maintained". The current regulation, 9 C.F.R. § 318.10 (1977), provides: "All parts of the pork muscle tissue shall be heated to a temperature not lower than 137° F., and the method used shall be one known to insure such a result." 9 C.F.R. § 318.10(c)(1) (1977). The same regulation was codified as early as 1949 as 9 C.F.R. § 18.10(c)(1) (1949). These regulations have been assumed for years to have been properly promulgated by the Secretary of Agriculture under the statutory authority given to him.41
Lastly, a holding that the regulation of smoked fish against the hazards of botulism is invalid for lack of authority would probably invalidate, to the extent that our ruling would be followed, the regulations concerning the purity of raw materials before their entry into the manufacturing process in 21 C.F.R. Part 113 (1977) (inspection of incoming raw materials for microbiological contamination before thermal processing of low-acid foods packed in hermetically sealed containers), in 21 C.F.R. Part 118 (1977) (pasteurization of milk and egg products to destroy Salmonella micro-organisms before use of the products in cacao products and confectionery), and 21 C.F.R. Part 129 (1977) (product water supply for processing and bottling of bottled drinking water must be of a safe, sanitary quality when it enters the process).42
The public interest will not permit invalidation simply on the basis of a lack of delegated statutory authority in this case. A gap in public health protection should not be created in the absence of a compelling reading based upon the utter absence of any statutory authority, even read expansively. Here we find no congressional history on the specific issue involved, and hence no impediment to the broader reading based on general purpose. We believe, nevertheless, that it would be in the public interest for Congress to consider in the light of existing knowledge, a legislative scheme for administrative regulation of the processing of food where hazard from micro-organisms in food in its natural state may require affirmative procedures of sterilization. This would entail, as well, a decision on the type of rulemaking procedure Congress thinks fit to impose.43
Appellants contend that there is an inadequate administrative record upon which to predicate judicial review, and that the failure to disclose to interested persons the factual material upon which the agency was relying vitiates the element of fairness which is essential to any kind of administrative action. Moreover, they argue that the "concise general statement of . . .  basis and purpose" by the Commissioner was inadequate. 5 U.S.C. § 553.45
The question of what is an adequate "record" in informal rulemaking has engaged the attention of commentators for several years. The extent of the administrative record required for judicial review of informal rulemaking is largely a function of the scope of judicial review. Even when the standard of review is whether the promulgation of the rule was "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law," as specified in 5 U.S.C. § 706(2)(A), judicial review must nevertheless, be based on the "whole record" (id.). Adequate review of a determination requires an adequate record, if the review is to be meaningful. Davis, Administrative Law in the Seventies, supra, at 669-71. What will constitute an adequate record for meaningful review may vary with the nature of the admininistrative action to be reviewed. Friendly, "Some Kind of Hearing," 123 U.Pa.L.Rev. 1267, 1291-92 (1975). Review must be based on the whole record even when the judgment is one of policy, except that findings of fact such as would be required in an adjudicatory proceeding or in a formal "on the record" hearing for rulemaking need not be made. Overton Park, supra, 401 U.S. at 416-18, 91 S.Ct. at 823-25 (1971). Though the action was informal, without an evidentiary record, the review must be "thorough, probing, [and] in depth". Id., 401 U.S. at 415, 91 S.Ct. 814. See Scalia & Goodman, Procedural Aspects of the Consumer Product Safety Act, 20 U.C.L.A.L.Rev. 899, 934-35 (1973).46
This raises several questions regarding the informal rulemaking procedure followed here: (1) What record does a reviewing court look to? (2) How much of what the agency relied on should have been disclosed to interested persons? (3) To what extent must the agency respond to criticism that is material?47
With respect to the content of the administrative "record," the Supreme Court has told us that in informal rulemaking, "the focal point for judicial review should be the administrative record already in existence, not some new record made initially in the reviewing court." See Camp v. Pitts, 411 U.S. 138, 142, 93 S.Ct. 1241, 1244, 36 L.Ed.2d 106 (1973).49
No contemporaneous record was made or certified. When, during the enforcement action, the basis for the regulation was sought through pretrial discovery, the record was created by searching the files of the FDA and the memories of those who participated in the process of rulemaking. This resulted in what became Exhibit D at the trial of the injunction action. Exhibit D consists of (1) Tab A containing the comments received from outside parties during the administrative "notice-and-comment"  proceeding and (2) Tabs B through L consisting of scientific data and the like upon which the Commissioner now says he relied but which were not made known to the interested parties.50
Appellants object to the exclusion of evidence in the District Court "aimed directly at showing that the scientific evidence relied upon by the FDA was inaccurate and not based upon a realistic appraisal of the true facts. Appellants attempted to introduce scientific evidence to demonstrate that in fixing the processing parameters FDA relied upon tests in which ground fish were injected with many millions of botulism [sic] spores and then tested for outgrowth at various processing levels whereas the spore levels in nature are far less and outgrowth would have been prevented by far less stringent processing parameters." (Br. p. 33). The District Court properly excluded the evidence.51
In an enforcement action, we must rely exclusively on the record made before the agency to determine the validity of the regulation. The exception to the exclusivity of that record is that "there may be independent judicial fact-finding when issues that were not before the agency are raised in a proceeding to enforce non-adjudicatory agency action." Overton Park, supra, 401 U.S. at 415, 91 S.Ct. at 823 (1971). (Emphasis added.)52
Though this is an enforcement proceeding and the question is close, we think that the "issues" were fairly before the agency and hence that de novo evidence was properly excluded by Judge Dooling. Camp v. Pitts, supra. Our concern is, rather, with the manner in which the agency treated the issues tendered.53
The key issues were (1) whether, in the light of the rather scant history of botulism in whitefish, that species should have been considered separately rather than included in a general regulation which failed to distinguish species from species; (2) whether the application of the proposed T-T-S requirements to smoked whitefish made the whitefish commercially unsaleable; and (3) whether the agency recognized that prospect, but nevertheless decided that the public health needs should prevail even if that meant commercial death for the whitefish industry. The procedural issues were whether, in the light of these key questions, the agency procedure was inadequate because (i) it failed to disclose to interested parties the scientific data and the methodology upon which it relied; and (ii) because it failed utterly to address itself to the pertinent question of commercial feasibility.55
The history of botulism occurrence in whitefish, as established in the trial record, which we must assume was available to the FDA in 1970, is as follows. Between 1899 and 1964 there were only eight cases of botulism reported as attributable to hot-smoked whitefish. In all eight instances, vacuum-packed whitefish was involved. All of the eight cases occurred in 1960 and 1963. The industry has abandoned vacuum-packing, and there has not been a single case of botulism associated with commercially prepared whitefish since 1963, though 2,750,000 pounds of whitefish are processed annually. Thus, in the seven-year period from 1964 through 1970, 17.25 million pounds of whitefish have been commercially processed in the United States without a single reported case of botulism. The evidence  also disclosed that defendant Nova Scotia has been in business some 56 years, and that there has never been a case of botulism illness from the whitefish processed by it.58
Interested parties were not informed of the scientific data, or at least of a selection of such data deemed important by the agency, so that comments could be addressed to the data. Appellants argue that unless the scientific data relied upon by the agency are spread upon the public records, criticism of the methodology used or the meaning to be inferred from the data is rendered impossible.61
We agree with appellants in this case, for although we recognize that an agency may resort to its own expertise outside the record in an informal rulemaking procedure, we do not believe that when the pertinent research material is readily available and the agency has no special expertise on the precise parameters involved, there is any reason to conceal the scientific data relied upon from the interested parties. As Judge Leventhal said in Portland Cement Ass'n v. Ruckelhaus, 158 U.S.App.D.C. 308, 326, 486 F.2d 375, 393 (1973): "It is not consonant with the purpose of a rulemaking proceeding to promulgate rules on the basis of inadequate data, or on data that [in] critical degree, is known only to the agency." (Emphasis added.) This is not a case where the agency methodology was based on material supplied by the interested parties themselves. Cf. International Harvester Co. v. Ruckelhaus, 155 U.S.App.D.C. 411, 428, 478 F.2d 615, 632 (1973). Here all the scientific research was collected by the agency, and none of it was disclosed to interested parties as the material upon which the proposed rule would be fashioned. Nor was an articulate effort made to connect the scientific requirements to available technology that would make commercial survival possible, though the burden of proof was on the agency. This required it to "bear a burden of adducing a reasoned presentation supporting the reliability of its methodology." International Harvester, supra, 155 U.S.App.D.C. at 439, 478 F.2d at 643.62
Though a reviewing court will not match submission against counter-submission to decide whether the agency was correct in its conclusion on scientific matters (unless that conclusion is arbitrary), it will consider whether the agency has taken account of all "relevant factors and whether there has been a clear error of judgment." Overton Park, supra, 401 U.S. at 415-16, 91 S.Ct. at 823-24; Appalachian Power Co. v. Environmental Protection Agency, 477 F.2d 495, 507 (4th Cir. 1973). In this circuit we have said that "it is `arbitrary or capricious' for an agency not to take into account all relevant factors in making its determination." Hanly v. Mitchell, 460 F.2d 640, 648 (2d Cir.), cert. denied, 409 U.S. 990, 93 S.Ct. 313, 34 L.Ed.2d 256 (1972) (an enforcement action under NEPA).63
If the failure to notify interested persons of the scientific research upon which the agency was relying actually prevented the presentation of relevant comment, the agency may be held not to have considered all "the relevant factors." We can think of no sound reasons for secrecy or reluctance to expose to public view (with an exception for trade secrets or national security) the ingredients of the deliberative process. Cf. Mobil Oil Corp. v. FPC, 157 U.S.App.D.C. 235, 256-58, 483 F.2d 1238, 1259-61 (1973). Indeed, the FDA's own regulations now specifically require that every notice of proposed rulemaking contain "references to all data and information on which the Commissioner relies for the proposal (copies or a full list of which shall be a part of the administrative file on the matter.  . .)." 21 C.F.R. § 10.40(b)(1) (1977). And this is, undoubtedly, the trend. See, e.g., National Nutritional Foods v. Weinberger, 512 F.2d 688 (2d Cir.), cert. denied, 423 U.S. 827, 96 S.Ct. 44, 46 L.Ed.2d 44 (1975).64
We think that the scientific data should have been disclosed to focus on the proper interpretation of "insanitary conditions." When the basis for a proposed rule is a scientific decision, the scientific material which is believed to support the rule should be exposed to the view of interested parties for their comment. One cannot ask for comment on a scientific paper without allowing the participants to read the paper. Scientific research is sometimes rejected for diverse inadequacies of methodology; and statistical results are sometimes rebutted because of a lack of adequate gathering technique or of supportable extrapolation. Such is the stuff of scientific debate. To suppress meaningful comment by failure to disclose the basic data relied upon is akin to rejecting comment altogether. For unless there is common ground, the comments are unlikely to be of a quality that might impress a careful agency. The inadequacy of comment in turn leads in the direction of arbitrary decision-making. We do not speak of findings of fact, for such are not technically required in the informal rulemaking procedures. We speak rather of what the agency should make known so as to elicit comments that probe the fundamentals. Informal rulemaking does not lend itself to a rigid pattern. Especially, in the circumstance of our broad reading of statutory authority in support of the agency, we conclude that the failure to disclose to interested persons the scientific data upon which the FDA relied was procedurally erroneous. Moreover, the burden was upon the agency to articulate rationally why the rule should apply to a large and diverse class, with the same T-T-S parameters made applicable to all species. Cf. Associated Industries of N.Y.S., Inc. v. U.S. Dept. of Labor, 487 F.2d 342, 352-53 (2d Cir. 1973). And cf. Industrial Union Dept. AFL-CIO v. Hodgson, 162 U.S.App.D.C. 331, 499 F.2d 467 (1974).65
Appellants additionally attack the "concise general statement" required by APA, 5 U.S.C. § 553, as inadequate. We think that, in the circumstances, it was less than adequate. It is not in keeping with the rational process to leave vital questions, raised by comments which are of cogent materiality, completely unanswered. The agencies certainly have a good deal of discretion in expressing the basis of a rule, but the agencies do not have quite the prerogative of obscurantism reserved to legislatures. "Congress did not purport to transfer its legislative power to the unbounded discretion of the regulatory body." F.C.C. v. RCA Communications, Inc., 346 U.S. 86, 90, 73 S.Ct. 998, 1002, 97 L.Ed. 1470 (1953) (Frankfurter, J.). As was said in Environmental Defense Fund, Inc. v. EPA, 150 U.S.App.D.C. 348, 371, 465 F.2d 528, 540-51 (1972): "We cannot discharge our role adequately unless we hold EPA to a high standard of articulation. Kennecott Copper Corp. v. EPA, . . . 149 U.S.App.D.C. 231, 462 F.2d 846 (1972)."67
The test of adequacy of the "concise general statement" was expressed by Judge McGowan in the following terms:68
"We do not expect the agency to discuss every item of fact or opinion included in the submissions made to it in informal rulemaking. We do expect that, if the judicial review which Congress has thought it important to provide is to be meaningful, the `concise general statement of . . . basis and purpose' mandated by Section 4 will enable us to see what major issues of policy were ventilated by the informal proceedings and why the agency reacted to them as it did." Automotive Parts & Accessories Ass'n v. Boyd, 132 U.S.App.D.C. 200, 208, 407 F.2d 330, 338 (1968).
And Judge Friendly has noted that "[i]n a case where a proposed standard under  OSHA [Occupational Safety and Health Act] has been opposed on grounds as substantial as those presented here, the Department has the burden of offering some reasoned explanation." Associated Industries of New York State, Inc. v. U.S. Department of Labor, supra, 487 F.2d at 352 (emphasis in original).70
The Secretary was squarely faced with the question whether it was necessary to formulate a rule with specific parameters that applied to all species of fish, and particularly whether lower temperatures with the addition of nitrite and salt would not be sufficient. Though this alternative was suggested by an agency of the federal government, its suggestion, though acknowledged, was never answered.71
Moreover, the comment that to apply the proposed T-T-S requirements to whitefish would destroy the commercial product was neither discussed nor answered. We think that to sanction silence in the face of such vital questions would be to make the statutory requirement of a "concise general statement" less than an adequate safeguard against arbitrary decision-making.72
We cannot improve on the statement of the District of Columbia Circuit in Industrial Union Dep't, AFL-CIO v. Hodgson, 162 U.S.App.D.C. 331, 339, 499 F.2d 467, 475 (1974).73
"What we are entitled to at all events is a careful identification by the Secretary, when his proposed standards are challenged, of the reasons why he chooses to follow one course rather than another. Where that choice purports to be based on the existence of certain determinable facts, the Secretary must, in form as well as in substance, find those facts from evidence in the record. By the same token, when the Secretary is obliged to make policy judgments where no factual certainties exist or where facts alone do not provide the answer, he should so state and go on to identify the considerations he found to be persuasive."
One may recognize that even commercial infeasibility cannot stand in the way of an overwhelming public interest. Yet the administrative process should disclose, at least, whether the proposed regulation is considered to be commercially feasible, or whether other considerations prevail even if commercial infeasibility is acknowledged. This kind of forthright disclosure and basic statement was lacking in the formulation of the T-T-S standard made applicable to whitefish. It is easy enough for an administrator to ban everything. In the regulation of food processing, the worldwide need for food also must be taken into account in formulating measures taken for the protection of health. In the light of the history of smoked whitefish to which we have referred, we find no articulate balancing here sufficient to make the procedure followed less than arbitrary.75
After seven years of relative inaction, the FDA has apparently not reviewed the T-T-S regulations in the light of present scientific knowledge and experience. In the absence of a new statutory directive by Congress regarding control of micro-organisms, which we hope will be worthy of its consideration, we think that the T-T-S standards should be reviewed again by the FDA.76
We cannot, on this appeal, remand to the agency to allow further comments by interested parties, addressed to the scientific data now disclosed at the trial below. We hold in this enforcement proceeding, therefore, that the regulation, as it affects non-vacuum-packed hot-smoked whitefish, was promulgated in an arbitrary manner and is invalid.77
When the District Court held the regulation to be valid, it properly exercised its discretion to grant the injunction. In view of our conclusion to the contrary, we must reverse the grant of the injunction and direct that the complaint be dismissed.78
 Hon. M. Joseph Blumenfeld, U. S. District Judge for the District of Connecticut, sitting by designation.79
 These regulations were codified previously in 21 C.F.R. Part 128a (1976). Defendant-appellants are the Nova Scotia Food Products Corp. ("Nova Scotia"), David Sklar, its president, and Emanuel Sklar, its vice-president and treasurer. Intervenor-appellant is the National Fisheries Institute, a trade association.80
 We shall for convenience refer to the Code Sections of 21 U.S.C. throughout rather than to the Sections of the Act in view of the number of amendments of the original statute.81
 We do not stress sanitation requirements in the regulation other than the heating and brining requirements because it is agreed that compliance with the other requirements does not prevent production of a commercially marketable smoked whitefish.82
 Appellants contend further that the regulation should be treated as merely interpretive rather than substantive and hence as an insufficient basis for the granting of injunctive relief. They also argue that the injunction is harsh and oppressive because it would destroy the whitefish business of appellants, and is unfair as selective enforcement against appellants alone.83
 The final regulations are codified at 21 C.F.R. Part 122 (1977).84
 "The food additive of sodium nitrite may be safely used in conjunction with salt (NaCl) to aid in inhibiting the outgrowth and toxin formation from Clostridium botulinum type E in the commercial processing of smoked chub . .: . . . [each fish shall be heated to] at least 160° F. for a minimum of 30 minutes [provided that it has not less than 3.5% of salt and a prescribed sodium nitrite content]." Note that whitefish at this temperature would require not less than 5% salt content under the regulation, 21 C.F.R. Part 122 (1977).85
 The administrative record, discussed in Part II, was introduced in evidence in the District Court as Exhibit D.86
 "[T]he `plain meaning' doctrine has always been subservient to a truly discernible legislative purpose however discerned". District of Columbia v. Orleans, 132 U.S.App.D.C. 139, 141, 406 F.2d 957, 959 (1968).87
 "The reviewing court shall —88
(2) hold unlawful and set aside agency action, findings and conclusions found to be —89
(C) in excess of statutory jurisdiction, authority, or limitations, or short of statutory right".90
 In December, 1972, Chief Counsel Hutt, speaking to the Annual Educational Conference of the Food and Drug Law Institute said, "[T]he Act must be regarded as a constitution." "[T]he fact that Congress simply has not considered or spoken on a particular issue certainly is no bar to the [FDA] exerting initiative and leadership in the public interest." 28 Food Drug Cosmetic Law Journal 177, 178-79 (March 1973). For a reply, see H. Thomas Austern, id. at 189 (March 1973). We do not take sides on the issue tendered, but we think Mr. Hutt's language to be conscious hyperbole. The test is not "initiative" but whether delegation may be fairly inferred from the general purpose.91
 Even under the standard of "arbitrary, capricious, an abuse of discretion or otherwise not in accordance with law," § 706(2)(A), generally used in informal rulemaking review, "the court shall review the whole record . . . and due account shall be taken of the rule of prejudicial error." § 706 (Emphasis added.) See Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 419, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971); Chicago v. FPC, 147 U.S.App.D.C. 312, 323, 458 F.2d 731, 744 (1971), cert. denied, 405 U.S. 1074, 92 S.Ct. 1495, 31 L.Ed.2d 808 (1972).92
 See, e. g., Hamilton, Procedures for the Adoption of Rules of General Applicability, 60 Cal.L.Rev. 1276, 1333-36(1972); Verkuil Judicial Review of Informal Rulemaking, 60 Va.L. Rev. 185 (1974); Pedersen, Formal Records and Informal Rulemaking, 85 Yale L.J. 38, 45-51 (1975). And see K. Davis, Administrative Law of the Seventies § 29.01-6 (1976).93
 A practice developed in the early years of the APA of not making a formal contemporaneous record, but rather, when challenged, to put together a historical record of what had been available for agency consideration at the time the regulation was promulgated. "One can conclude that `record' now means whatever the agency produces on review." Verkuil, supra, 60 Va.L.Rev. at 204. And see Deutsche Lufthansa, A.G. v. CAB, 156 U.S.App.D.C. 191, 479 F.2d 912 (1973) (record unchallenged). Professor Davis in a balanced review, has stated: "When the facts are of central importance and might be challenged, parties adversely affected by them should have a chance to respond to them. Clearly, whatever "factual information the agency has considered should be a part of the record for judicial review." K. Davis, supra, § 29.01-6, pp. 672-73.94
 The authorities cited by appellants purporting to show the inadequacy of the record involve administrative actions required to be taken only on the basis of evidentiary hearings, e. g., Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 288 n. 4, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974), and Hamilton, Administrative Rulemaking, supra, 60 Cal.L.Rev. at 1281-82; Williams v. Robinson, 139 U.S.App.D.C. 204, 432 F.2d 637 (1970); or ratemaking for utilities by state agencies in which no evidence to support the agency's decision was produced even on appeal. Ohio Bell Tel. Co. v. PUC, 301 U.S. 292, 300, 57 S.Ct. 724, 81 L.Ed. 1093 (1937).95
 We recognize the problem posed by Judge Leventhal in International Harvester, supra, that a proceeding might never end if such submission required a reply ad infinitum, ibid. Here the exposure of the scientific research relied on simply would have required a single round of comment addressed thereto.96
 And see Judge Friendly's discussion of recent developments in rulemaking and judicial review in Friendly, "Some Kind of Hearing," 123 U.Pa.L.Rev. 1267, 1305-15 (1975).97
 There informal agency procedure was followed, though with a standard of review traditionally conceived of as suited to formal adjudication or rulemaking.
463 U.S. 29 (1983)2
Supreme Court of United States.
Argued April 26, 1983.
Decided June 24, 1983.
CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE DISTRICT OF COLUMBIA CIRCUIT.5
 Solicitor General Lee argued the cause for petitioners in No. 82-398. With him on the briefs were Assistant Attorney General McGrath, Deputy Solicitor General Geller, Edwin S. Kneedler, Robert E. Kopp, Michael F. Hertz, Frank Berndt, David W. Allen, Enid Rubenstein, and Eileen T. Leahy. Lloyd N. Cutler argued the cause for petitioners in No. 82-354. With him on the briefs were John H. Pickering, William R. Perlik, Andrew B. Weissman, William R. Richardson, Jr., Milton D. Andrews, Lance E. Tunick, William H. Crabtree, Edward P. Good, Henry R. Nolte, Jr., Otis M. Smith, Charles R. Sharp, and William L. Weber, Jr. Raymond M. Momboisse, Sam Kazman, and Ronald A. Zumbrun filed briefs for petitioners in No. 82-355.6
James F. Fitzpatrick argued the cause for respondents in all cases. With him on the brief for respondents State Farm Mutual Automobile Insurance Co., et al. were Michael N. Sohn, John M. Quinn, and Merrick B. Garland. Robert Abrams, Attorney General of New York, Robert S. Hammer, Assistant Attorney General, Peter H. Schiff, Martin Minkowitz, and Milton L. Freedman filed a brief for respondent Superintendent of Insurance of the State of New York. Raymond J. Rasenberger, Lawrence C. Merthan, Jerry W. Cox, and Lowell R. Beck filed a brief for respondents National Association of Independent Insurers et al.7
The development of the automobile gave Americans unprecedented freedom to travel, but exacted a high price for  enhanced mobility. Since 1929, motor vehicles have been the leading cause of accidental deaths and injuries in the United States. In 1982, 46,300 Americans died in motor vehicle accidents and hundreds of thousands more were maimed and injured. While a consensus exists that the current loss of life on our highways is unacceptably high, improving safety does not admit to easy solution. In 1966, Congress decided that at least part of the answer lies in improving the design and safety features of the vehicle itself. But much of the technology for building safer cars was undeveloped or untested. Before changes in automobile design could be mandated, the effectiveness of these changes had to be studied, their costs examined, and public acceptance considered. This task called for considerable expertise and Congress responded by enacting the National Traffic and Motor Vehicle Safety Act of 1966 (Act), 80 Stat. 718, as amended, 15 U. S. C. § 1381 et seq. (1976 ed. and Supp. V). The Act, created for the purpose of "reduc[ing] traffic accidents and deaths and injuries to persons resulting from traffic accidents," 15 U. S. C. § 1381, directs the Secretary of Transportation or his delegate to issue motor vehicle safety standards that "shall be practicable, shall meet the need for motor vehicle safety, and shall be stated in objective terms." 15 U. S. C. § 1392(a) (1976 ed., Supp. V). In issuing these standards, the Secretary is directed to consider "relevant available motor vehicle safety data," whether the proposed standard "is reasonable, practicable and appropriate" for the particular type of motor vehicle, and the "extent to which  such standards will contribute to carrying out the purposes" of the Act. 15 U. S. C. §§ 1392(f)(1), (3), (4).9
The Act also authorizes judicial review under the provisions of the Administrative Procedure Act (APA), 5 U. S. C. § 706, of all "orders establishing, amending, or revoking a Federal motor vehicle safety standard," 15 U. S. C. § 1392(b). Under this authority, we review today whether NHTSA acted arbitrarily and capriciously in revoking the requirement in Motor Vehicle Safety Standard 208 that new motor vehicles produced after September 1982 be equipped with passive restraints to protect the safety of the occupants of the vehicle in the event of a collision. Briefly summarized, we hold that the agency failed to present an adequate basis and explanation for rescinding the passive restraint requirement and that the agency must either consider the matter further or adhere to or amend Standard 208 along lines which its analysis supports.10
The regulation whose rescission is at issue bears a complex and convoluted history. Over the course of approximately 60 rulemaking notices, the requirement has been imposed, amended, rescinded, reimposed, and now rescinded again.12
As originally issued by the Department of Transportation in 1967, Standard 208 simply required the installation of seatbelts in all automobiles. 32 Fed. Reg. 2415. It soon became apparent that the level of seatbelt use was too low to reduce traffic injuries to an acceptable level. The Department therefore began consideration of "passive occupant restraint systems" — devices that do not depend for their effectiveness  upon any action taken by the occupant except that necessary to operate the vehicle. Two types of automatic crash protection emerged: automatic seatbelts and airbags. The automatic seatbelt is a traditional safety belt, which when fastened to the interior of the door remains attached without impeding entry or exit from the vehicle, and deploys automatically without any action on the part of the passenger. The airbag is an inflatable device concealed in the dashboard and steering column. It automatically inflates when a sensor indicates that deceleration forces from an accident have exceeded a preset minimum, then rapidly deflates to dissipate those forces. The lifesaving potential of these devices was immediately recognized, and in 1977, after substantial on-the-road experience with both devices, it was estimated by NHTSA that passive restraints could prevent approximately 12,000 deaths and over 100,000 serious injuries annually. 42 Fed. Reg. 34298.13
In 1969, the Department formally proposed a standard requiring the installation of passive restraints, 34 Fed. Reg. 11148, thereby commencing a lengthy series of proceedings. In 1970, the agency revised Standard 208 to include passive protection requirements, 35 Fed. Reg. 16927, and in 1972, the agency amended the Standard to require full passive protection for all front seat occupants of vehicles manufactured after August 15, 1975. 37 Fed. Reg. 3911. In the interim, vehicles built between August 1973 and August 1975 were to carry either passive restraints or lap and shoulder belts coupled with an "ignition interlock" that would prevent starting the vehicle if the belts were not connected. On review, the  agency's decision to require passive restraints was found to be supported by "substantial evidence" and upheld. Chrysler Corp. v. Department of Transportation, 472 F. 2d 659 (CA6 1972).14
In preparing for the upcoming model year, most car makers chose the "ignition interlock" option, a decision which was highly unpopular, and led Congress to amend the Act to prohibit a motor vehicle safety standard from requiring or permitting compliance by means of an ignition interlock or a continuous buzzer designed to indicate that safety belts were not in use. Motor Vehicle and Schoolbus Safety Amendments of 1974, Pub. L. 93-492, § 109, 88 Stat. 1482, 15 U. S. C. § 1410b(b). The 1974 Amendments also provided that any safety standard that could be satisfied by a system other than seatbelts would have to be submitted to Congress where it could be vetoed by concurrent resolution of both Houses. 15 U. S. C. § 1410b(b)(2).15
The effective date for mandatory passive restraint systems was extended for a year until August 31, 1976. 40 Fed. Reg. 16217 (1975); id., at 33977. But in June 1976, Secretary of Transportation William T. Coleman, Jr., initiated a new rulemaking on the issue, 41 Fed. Reg. 24070. After hearing testimony and reviewing written comments, Coleman extended the optional alternatives indefinitely and suspended the passive restraint requirement. Although he found passive  restraints technologically and economically feasible, the Secretary based his decision on the expectation that there would be widespread public resistance to the new systems. He instead proposed a demonstration project involving up to 500,000 cars installed with passive restraints, in order to smooth the way for public acceptance of mandatory passive restraints at a later date. Department of Transportation, The Secretary's Decision Concerning Motor Vehicle Occupant Crash Protection (Dec. 6, 1976), App. 2068.16
Coleman's successor as Secretary of Transportation disagreed. Within months of assuming office, Secretary Brock Adams decided that the demonstration project was unnecessary. He issued a new mandatory passive restraint regulation, known as Modified Standard 208. 42 Fed. Reg. 34289 (1977); 49 CFR § 571.208 (1978). The Modified Standard mandated the phasing in of passive restraints beginning with large cars in model year 1982 and extending to all cars by model year 1984. The two principal systems that would satisfy the Standard were airbags and passive belts; the choice of which system to install was left to the manufacturers. In Pacific Legal Foundation v. Department of Transportation, 193 U. S. App. D. C. 184, 593 F. 2d 1338, cert. denied, 444 U. S. 830 (1979), the Court of Appeals upheld Modified Standard 208 as a rational, nonarbitrary regulation consistent with the agency's mandate under the Act. The Standard also survived scrutiny by Congress, which did not exercise its authority under the legislative veto provision of the 1974 Amendments.17
Over the next several years, the automobile industry geared up to comply with Modified Standard 208. As late as July 1980, NHTSA reported:18
 "On the road experience in thousands of vehicles equipped with air bags and automatic safety belts has confirmed agency estimates of the life-saving and injury-preventing benefits of such systems. When all cars are equipped with automatic crash protection systems, each year an estimated 9,000 more lives will be saved, and tens of thousands of serious injuries will be prevented." NHTSA, Automobile Occupant Crash Protection, Progress Report No. 3, p. 4; App. in No. 81-2220 (CADC), p. 1627 (hereinafter App.).
In February 1981, however, Secretary of Transportation Andrew Lewis reopened the rulemaking due to changed economic circumstances and, in particular, the difficulties of the automobile industry. 46 Fed. Reg. 12033. Two months later, the agency ordered a one-year delay in the application of the Standard to large cars, extending the deadline to September 1982, id., at 21172, and at the same time, proposed the possible rescission of the entire Standard. Id., at 21205. After receiving written comments and holding public hearings, NHTSA issued a final rule (Notice 25) that rescinded the passive restraint requirement contained in Modified Standard 208.20
In a statement explaining the rescission, NHTSA maintained that it was no longer able to find, as it had in 1977, that the automatic restraint requirement would produce significant safety benefits. Notice 25, id., at 53419. This judgment reflected not a change of opinion on the effectiveness of the technology, but a change in plans by the automobile industry. In 1977, the agency had assumed that airbags would be installed in 60% of all new cars and automatic seatbelts in 40%. By 1981 it became apparent that automobile manufacturers planned to install the automatic seatbelts in approximately 99% of the new cars. For this reason, the lifesaving potential of airbags would not be realized. Moreover, it now appeared that the overwhelming majority of passive belts  planned to be installed by manufacturers could be detached easily and left that way permanently. Passive belts, once detached, then required "the same type of affirmative action that is the stumbling block to obtaining high usage levels of manual belts." Id., at 53421. For this reason, the agency concluded that there was no longer a basis for reliably predicting that the Standard would lead to any significant increased usage of restraints at all.22
In view of the possibly minimal safety benefits, the automatic restraint requirement no longer was reasonable or practicable in the agency's view. The requirement would require approximately $1 billion to implement and the agency did not believe it would be reasonable to impose such substantial costs on manufacturers and consumers without more adequate assurance that sufficient safety benefits would accrue. In addition, NHTSA concluded that automatic restraints might have an adverse effect on the public's attitude toward safety. Given the high expense and limited benefits of detachable belts, NHTSA feared that many consumers would regard the Standard as an instance of ineffective regulation, adversely affecting the public's view of safety regulation and, in particular, "poisoning . . . popular sentiment toward efforts to improve occupant restraint systems in the future." Id., at 53424.23
State Farm Mutual Automobile Insurance Co. and the National Association of Independent Insurers filed petitions for review of NHTSA's rescission of the passive restraint Standard. The United States Court of Appeals for the District of Columbia Circuit held that the agency's rescission of the passive restraint requirement was arbitrary and capricious. 220 U. S. App. D. C. 170, 680 F. 2d 206 (1982). While observing that rescission is not unrelated to an agency's refusal to take action in the first instance, the court concluded that, in this case, NHTSA's discretion to rescind the passive restraint requirement had been restricted by various forms of congressional "reaction" to the passive restraint issue. It then  proceeded to find that the rescission of Standard 208 was arbitrary and capricious for three reasons. First, the court found insufficient as a basis for rescission NHTSA's conclusion that it could not reliably predict an increase in belt usage under the Standard. The court held that there was insufficient evidence in the record to sustain NHTSA's position on this issue, and that, "only a well justified refusal to seek more evidence could render rescission non-arbitrary." Id., at 196, 680 F. 2d, at 232. Second, a majority of the panel concluded that NHTSA inadequately considered the possibility of requiring manufacturers to install nondetachable rather than detachable passive belts. Third, the majority found that the agency acted arbitrarily and capriciously by failing to give any consideration whatever to requiring compliance with Modified Standard 208 by the installation of airbags.24
The court allowed NHTSA 30 days in which to submit a schedule for "resolving the questions raised in th[e] opinion." Id., at 206, 680 F. 2d, at 242. Subsequently, the agency filed a Notice of Proposed Supplemental Rulemaking setting forth a schedule for complying with the court's mandate. On August 4, 1982, the Court of Appeals issued an order staying the compliance date for the passive restraint requirement until September 1, 1983, and requested NHTSA to inform the court whether that compliance date was achievable. NHTSA informed the court on October 1, 1982, that based on representations by manufacturers, it did not appear that practicable compliance could be achieved before September 1985. On November 8, 1982, we granted certiorari, 459 U. S. 987, and on November 18, the Court of Appeals entered an order recalling its mandate.25
Unlike the Court of Appeals, we do not find the appropriate scope of judicial review to be the "most troublesome  question" in these cases. Both the Act and the 1974 Amendments concerning occupant crash protection standards indicate that motor vehicle safety standards are to be promulgated under the informal rulemaking procedures of the Administrative Procedure Act. 5 U. S. C. § 553. The agency's action in promulgating such standards therefore may be set aside if found to be "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U. S. C. § 706(2)(A); Citizens to Preserve Overton Park v. Volpe, 401 U. S. 402, 414 (1971); Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc., 419 U. S. 281 (1974). We believe that the rescission or modification of an occupant-protection standard is subject to the same test. Section 103(b) of the Act, 15 U. S. C. § 1392(b), states that the procedural and judicial review provisions of the Administrative Procedure Act "shall apply to all orders establishing, amending, or revoking a Federal motor vehicle safety standard," and suggests no difference in the scope of judicial review depending upon the nature of the agency's action.27
Petitioner Motor Vehicle Manufacturers Association (MVMA) disagrees, contending that the rescission of an agency rule should be judged by the same standard a court would use to judge an agency's refusal to promulgate a rule in the first place — a standard petitioner believes considerably narrower than the traditional arbitrary-and-capricious test. We reject this view. The Act expressly equates orders "revoking" and "establishing" safety standards; neither that Act nor the APA suggests that revocations are to be treated as refusals to promulgate standards. Petitioner's view would render meaningless Congress' authorization for judicial review of orders revoking safety rules. Moreover, the revocation of an extant regulation is substantially different than a failure to act. Revocation constitutes a reversal of the agency's former views as to the proper course. A "settled course of behavior embodies the agency's informed judgment that, by pursuing that course, it will carry out the policies  committed to it by Congress. There is, then, at least a presumption that those policies will be carried out best if the settled rule is adhered to." Atchison, T. & S. F. R. Co. v. Wichita Bd. of Trade, 412 U. S. 800, 807-808 (1973). Accordingly, an agency changing its course by rescinding a rule is obligated to supply a reasoned analysis for the change beyond that which may be required when an agency does not act in the first instance.28
In so holding, we fully recognize that "[r]egulatory agencies do not establish rules of conduct to last forever," American Trucking Assns., Inc. v. Atchison, T. & S. F. R. Co., 387 U. S. 397, 416 (1967), and that an agency must be given ample latitude to "adapt their rules and policies to the demands of changing circumstances." Permian Basin Area Rate Cases, 390 U. S. 747, 784 (1968). But the forces of change do not always or necessarily point in the direction of deregulation. In the abstract, there is no more reason to presume that changing circumstances require the rescission of prior action, instead of a revision in or even the extension of current regulation. If Congress established a presumption from which judicial review should start, that presumption — contrary to petitioners' views — is not against safety regulation, but against changes in current policy that are not justified by the rulemaking record. While the removal of a regulation may not entail the monetary expenditures and other costs of enacting a new standard, and, accordingly, it may be easier for an agency to justify a deregulatory action, the direction in which an agency chooses to move does not alter the standard of judicial review established by law.29
The Department of Transportation accepts the applicability of the "arbitrary and capricious" standard. It argues that under this standard, a reviewing court may not set aside an agency rule that is rational, based on consideration of the relevant factors, and within the scope of the authority delegated to the agency by the statute. We do not disagree with  this formulation. The scope of review under the "arbitrary and capricious" standard is narrow and a court is not to substitute its judgment for that of the agency. Nevertheless, the agency must examine the relevant data and articulate a satisfactory explanation for its action including a "rational connection between the facts found and the choice made." Burlington Truck Lines, Inc. v. United States, 371 U. S. 156, 168 (1962). In reviewing that explanation, we must "consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment." Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc., supra, at 285; Citizens to Preserve Overton Park v. Volpe, supra, at 416. Normally, an agency rule would be arbitrary and capricious if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise. The reviewing court should not attempt itself to make up for such deficiencies; we may not supply a reasoned basis for the agency's action that the agency itself has not given. SEC v. Chenery Corp., 332 U. S. 194, 196 (1947). We will, however, "uphold a decision of less than ideal clarity if the agency's path may reasonably be discerned." Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc., supra, at 286. See also Camp v. Pitts, 411 U. S. 138, 142-143 (1973) (per curiam). For purposes of these cases, it is also relevant that Congress required a record of the rulemaking proceedings to be compiled  and submitted to a reviewing court, 15 U. S. C. § 1394, and intended that agency findings under the Act would be supported by "substantial evidence on the record considered as a whole." S. Rep. No. 1301, 89th Cong., 2d Sess., 8 (1966); H. R. Rep. No. 1776, 89th Cong., 2d Sess., 21 (1966).30
The Court of Appeals correctly found that the arbitrary-and-capricious test applied to rescissions of prior agency regulations, but then erred in intensifying the scope of its review based upon its reading of legislative events. It held that congressional reaction to various versions of Standard 208 "raise[d] doubts" that NHTSA's rescission "necessarily demonstrates an effort to fulfill its statutory mandate," and therefore the agency was obligated to provide "increasingly clear and convincing reasons" for its action. 220 U. S. App. D. C., at 186, 193, 680 F. 2d, at 222, 229. Specifically, the Court of Appeals found significance in three legislative occurrences:32
"In 1974, Congress banned the ignition interlock but did not foreclose NHTSA's pursuit of a passive restraint standard. In 1977, Congress allowed the standard to take effect when neither of the concurrent resolutions needed for disapproval was passed. In 1980, a majority of each house indicated support for the concept of mandatory passive restraints and a majority of each house supported the unprecedented attempt to require some installation of airbags." Id., at 192, 680 F. 2d, at 228.
From these legislative acts and nonacts the Court of Appeals derived a "congressional commitment to the concept of automatic crash protection devices for vehicle occupants." Ibid.34
This path of analysis was misguided and the inferences it produced are questionable. It is noteworthy that in this Court respondent State Farm expressly agrees that the post-enactment legislative history of the Act does not heighten the  standard of review of NHTSA's actions. Brief for Respondent State Farm Mutual Automobile Insurance Co. 13. State Farm's concession is well taken for this Court has never suggested that the standard of review is enlarged or diminished by subsequent congressional action. While an agency's interpretation of a statute may be confirmed or ratified by subsequent congressional failure to change that interpretation, Bob Jones University v. United States, 461 U. S. 574, 599-602 (1983); Haig v. Agee, 453 U. S. 280, 291-300 (1981), in the cases before us, even an unequivocal ratification — short of statutory incorporation — of the passive restraint standard would not connote approval or disapproval of an agency's later decision to rescind the regulation. That decision remains subject to the arbitrary-and-capricious standard.35
That we should not be so quick to infer a congressional mandate for passive restraints is confirmed by examining the postenactment legislative events cited by the Court of Appeals. Even were we inclined to rely on inchoate legislative action, the inferences to be drawn fail to suggest that NHTSA acted improperly in rescinding Standard 208. First, in 1974 a mandatory passive restraint standard was technically not in effect, see n. 6, supra; Congress had no reason to foreclose that course. Moreover, one can hardly infer support for a mandatory standard from Congress' decision to provide that such a regulation would be subject to disapproval by resolutions of disapproval in both Houses. Similarly, no mandate can be divined from the tabling of resolutions of disapproval which were introduced in 1977. The failure of Congress to exercise its veto might reflect legislative deference to the agency's expertise and does not indicate that Congress would disapprove of the agency's action in 1981. And even if Congress favored the Standard in 1977, it — like NHTSA — may well reach a different judgment, given changed circumstances four years later. Finally, the Court of Appeals read too much into floor action on the 1980 authorization bill, a bill which was not enacted into law. Other  contemporaneous events could be read as showing equal congressional hostility to passive restraints.36
The ultimate question before us is whether NHTSA's rescission of the passive restraint requirement of Standard 208 was arbitrary and capricious. We conclude, as did the Court of Appeals, that it was. We also conclude, but for somewhat different reasons, that further consideration of the issue by the agency is therefore required. We deal separately with the rescission as it applies to airbags and as it applies to seatbelts.38
The first and most obvious reason for finding the rescission arbitrary and capricious is that NHTSA apparently gave no consideration whatever to modifying the Standard to require that airbag technology be utilized. Standard 208 sought to achieve automatic crash protection by requiring automobile manufacturers to install either of two passive restraint devices: airbags or automatic seatbelts. There was no suggestion in the long rulemaking process that led to Standard 208 that if only one of these options were feasible, no passive restraint standard should be promulgated. Indeed, the agency's original proposed Standard contemplated the installation of inflatable restraints in all cars. Automatic belts  were added as a means of complying with the Standard because they were believed to be as effective as airbags in achieving the goal of occupant crash protection. 36 Fed. Reg. 12859 (1971). At that time, the passive belt approved by the agency could not be detached. Only later, at a manufacturer's behest, did the agency approve of the detachability feature — and only after assurances that the feature would not compromise the safety benefits of the restraint. Although it was then foreseen that 60% of the new cars would contain airbags and 40% would have automatic seatbelts, the ratio between the two was not significant as long as the passive belt would also assure greater passenger safety.40
The agency has now determined that the detachable automatic belts will not attain anticipated safety benefits because so many individuals will detach the mechanism. Even if this conclusion were acceptable in its entirety, see infra, at 51-54, standing alone it would not justify any more than an amendment of Standard 208 to disallow compliance by means of the one technology which will not provide effective passenger protection. It does not cast doubt on the need for a passive restraint standard or upon the efficacy of airbag technology. In its most recent rulemaking, the agency again acknowledged the lifesaving potential of the airbag:41
 "The agency has no basis at this time for changing its earlier conclusions in 1976 and 1977 that basic air bag technology is sound and has been sufficiently demonstrated to be effective in those vehicles in current use . . . ." NHTSA Final Regulatory Impact Analysis (RIA) XI-4 (Oct. 1981), App. 264.
Given the effectiveness ascribed to airbag technology by the agency, the mandate of the Act to achieve traffic safety would suggest that the logical response to the faults of detachable seatbelts would be to require the installation of airbags. At the very least this alternative way of achieving the objectives of the Act should have been addressed and adequate reasons given for its abandonment. But the agency not only did not require compliance through airbags, it also did not even consider the possibility in its 1981 rulemaking. Not one sentence of its rulemaking statement discusses the airbags-only option. Because, as the Court of Appeals stated, "NHTSA's . . . analysis of airbags was nonexistent," 220 U. S. App. D. C., at 200, 680 F. 2d, at 236, what we said in Burlington Truck Lines, Inc. v. United States, 371 U. S., at 167, is apropos here:43
"There are no findings and no analysis here to justify the choice made, no indication of the basis on which the [agency] exercised its expert discretion. We are not prepared to and the Administrative Procedure Act will not permit us to accept such . . . practice. . . . Expert discretion is the lifeblood of the administrative process, but `unless we make the requirements for administrative action strict and demanding, expertise, the strength of modern government, can become a monster which rules with no practical limits on its discretion.' New York v. United States, 342 U. S. 882, 884 (dissenting opinion)" (footnote omitted).
We have frequently reiterated that an agency must cogently explain why it has exercised its discretion in a given manner,  Atchison, T. & S. F. R. Co. v. Wichita Bd. of Trade, 412 U. S., at 806; FTC v. Sperry & Hutchinson Co., 405 U. S. 233, 249 (1972); NLRB v. Metropolitan Life Ins. Co., 380 U. S. 438, 443 (1965); and we reaffirm this principle again today.45
The automobile industry has opted for the passive belt over the airbag, but surely it is not enough that the regulated industry has eschewed a given safety device. For nearly a decade, the automobile industry waged the regulatory equivalent of war against the airbag and lost — the inflatable restraint was proved sufficiently effective. Now the automobile industry has decided to employ a seatbelt system which will not meet the safety objectives of Standard 208. This hardly constitutes cause to revoke the Standard itself. Indeed, the Act was necessary because the industry was not sufficiently responsive to safety concerns. The Act intended that safety standards not depend on current technology and could be "technology-forcing" in the sense of inducing the development of superior safety design. See Chrysler Corp. v. Department of Transportation, 472 F. 2d, at 672-673. If, under the statute, the agency should not defer to the industry's failure to develop safer cars, which it surely should not do, a fortiori it may not revoke a safety standard which can be satisfied by current technology simply because the industry has opted for an ineffective seatbelt design.46
Although the agency did not address the mandatory airbag option and the Court of Appeals noted that "airbags seem to have none of the problems that NHTSA identified in passive seatbelts," 220 U. S. App. D. C., at 201, 680 F. 2d, at 237, petitioners recite a number of difficulties that they  believe would be posed by a mandatory airbag standard. These range from questions concerning the installation of airbags in small cars to that of adverse public reaction. But these are not the agency's reasons for rejecting a mandatory airbag standard. Not having discussed the possibility, the agency submitted no reasons at all. The short — and sufficient — answer to petitioners' submission is that the courts may not accept appellate counsel's post hoc rationalizations for agency action. Burlington Truck Lines, Inc. v. United States, 371 U. S., at 168. It is well established that an agency's action must be upheld, if at all, on the basis articulated by the agency itself. Ibid.; SEC v. Chenery Corp., 332 U. S., at 196; American Textile Mfrs. Institute, Inc. v. Donovan, 452 U. S. 490, 539 (1981).47
Petitioners also invoke our decision in Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc., 435 U. S. 519 (1978), as though it were a talisman under which any agency decision is by definition unimpeachable. Specifically, it is submitted that to require an agency to consider an airbags-only alternative is, in essence, to dictate to the agency the procedures it is to follow. Petitioners both misread Vermont Yankee and misconstrue the nature of the remand that is in order. In Vermont Yankee, we held that a court may not impose additional procedural requirements upon an agency. We do not require today any specific procedures  which NHTSA must follow. Nor do we broadly require an agency to consider all policy alternatives in reaching decision. It is true that rulemaking "cannot be found wanting simply because the agency failed to include every alternative device and thought conceivable by the mind of man . . . regardless of how uncommon or unknown that alternative may have been . . . ." Id., at 551. But the airbag is more than a policy alternative to the passive restraint Standard; it is a technological alternative within the ambit of the existing Standard. We hold only that given the judgment made in 1977 that airbags are an effective and cost-beneficial life-saving technology, the mandatory passive restraint rule may not be abandoned without any consideration whatsoever of an airbags-only requirement.48
Although the issue is closer, we also find that the agency was too quick to dismiss the safety benefits of automatic seatbelts. NHTSA's critical finding was that, in light of the industry's plans to install readily detachable passive belts, it could not reliably predict "even a 5 percentage point increase as the minimum level of expected usage increase." 46 Fed. Reg. 53423 (1981). The Court of Appeals rejected this finding because there is "not one iota" of evidence that Modified Standard 208 will fail to increase nationwide seatbelt use by at least 13 percentage points, the level of increased usage necessary for the Standard to justify its cost. Given the lack of probative evidence, the court held that "only a well justified refusal to seek more evidence could render rescission non-arbitrary." 220 U. S. App. D. C., at 196, 680 F. 2d, at 232.50
Petitioners object to this conclusion. In their view, "substantial uncertainty" that a regulation will accomplish its intended purpose is sufficient reason, without more, to rescind a regulation. We agree with petitioners that just as an agency reasonably may decline to issue a safety standard if it is uncertain about its efficacy, an agency may also revoke a  standard on the basis of serious uncertainties if supported by the record and reasonably explained. Rescission of the passive restraint requirement would not be arbitrary and capricious simply because there was no evidence in direct support of the agency's conclusion. It is not infrequent that the available data do not settle a regulatory issue, and the agency must then exercise its judgment in moving from the facts and probabilities on the record to a policy conclusion. Recognizing that policymaking in a complex society must account for uncertainty, however, does not imply that it is sufficient for an agency to merely recite the terms "substantial uncertainty" as a justification for its actions. As previously noted, the agency must explain the evidence which is available, and must offer a "rational connection between the facts found and the choice made." Burlington Truck Lines, Inc. v. United States, supra, at 168. Generally, one aspect of that explanation would be a justification for rescinding the regulation before engaging in a search for further evidence.51
In these cases, the agency's explanation for rescission of the passive restraint requirement is not sufficient to enable us to conclude that the rescission was the product of reasoned decisionmaking. To reach this conclusion, we do not upset the agency's view of the facts, but we do appreciate the limitations of this record in supporting the agency's decision. We start with the accepted ground that if used, seatbelts unquestionably would save many thousands of lives and would prevent tens of thousands of crippling injuries. Unlike recent regulatory decisions we have reviewed, Industrial Union Dept. v. American Petroleum Institute, 448 U. S. 607 (1980); American Textile Mfrs. Institute, Inc. v. Donovan, 452 U. S. 490 (1981), the safety benefits of wearing seatbelts are not in doubt, and it is not challenged that were those benefits to accrue, the monetary costs of implementing the Standard would be easily justified. We move next to the fact that there is no direct evidence in support of the agency's finding that detachable automatic belts cannot be predicted  to yield a substantial increase in usage. The empirical evidence on the record, consisting of surveys of drivers of automobiles equipped with passive belts, reveals more than a doubling of the usage rate experienced with manual belts. Much of the agency's rulemaking statement — and much of the controversy in these cases — centers on the conclusions that should be drawn from these studies. The agency maintained that the doubling of seatbelt usage in these studies could not be extrapolated to an across-the-board mandatory standard because the passive seatbelts were guarded by ignition interlocks and purchasers of the tested cars are somewhat atypical. Respondents insist these studies demonstrate that Modified Standard 208 will substantially increase seatbelt usage. We believe that it is within the agency's discretion to pass upon the generalizability of these field studies. This is precisely the type of issue which rests within the expertise of NHTSA, and upon which a reviewing court must be most hesitant to intrude.52
But accepting the agency's view of the field tests on passive restraints indicates only that there is no reliable real-world experience that usage rates will substantially increase. To be sure, NHTSA opines that "it cannot reliably predict even a 5 percentage point increase as the minimum level of  expected increased usage." Notice 25, 46 Fed. Reg. 53423 (1981). But this and other statements that passive belts will not yield substantial increases in seatbelt usage apparently take no account of the critical difference between detachable automatic belts and current manual belts. A detached passive belt does require an affirmative act to reconnect it, but — unlike a manual seatbelt — the passive belt, once reattached, will continue to function automatically unless again disconnected. Thus, inertia — a factor which the agency's own studies have found significant in explaining the current low usage rates for seatbelts — works in favor of, not against, use of the protective device. Since 20% to 50% of motorists currently wear seatbelts on some occasions, there would seem to be grounds to believe that seatbelt use by occasional users will be substantially increased by the detachable passive belts. Whether this is in fact the case is a matter for the agency to decide, but it must bring its expertise to bear on the question.53
The agency is correct to look at the costs as well as the benefits of Standard 208. The agency's conclusion that the incremental costs of the requirements were no longer reasonable was predicated on its prediction that the safety benefits of the regulation might be minimal. Specifically, the  agency's fears that the public may resent paying more for the automatic belt systems is expressly dependent on the assumption that detachable automatic belts will not produce more than "negligible safety benefits." Id., at 53424. When the agency reexamines its findings as to the likely increase in seatbelt usage, it must also reconsider its judgment of the reasonableness of the monetary and other costs associated with the Standard. In reaching its judgment, NHTSA should bear in mind that Congress intended safety to be the pre-eminent factor under the Act:54
"The Committee intends that safety shall be the overriding consideration in the issuance of standards under this bill. The Committee recognizes . . . that the Secretary will necessarily consider reasonableness of cost, feasibility and adequate leadtime." S. Rep. No. 1301, 89th Cong., 2d Sess., 6 (1966).
"In establishing standards the Secretary must conform to the requirement that the standard be practicable. This would require consideration of all relevant factors, including technological ability to achieve the goal of a particular standard as well as consideration of economic factors.
"Motor vehicle safety is the paramount purpose of this bill and each standard must be related thereto." H. R. Rep. No. 1776, 89th Cong., 2d Sess., 16 (1966).
The agency also failed to articulate a basis for not requiring nondetachable belts under Standard 208. It is argued that the concern of the agency with the easy detachability of the currently favored design would be readily solved by a continuous passive belt, which allows the occupant to "spool out" the belt and create the necessary slack for easy extrication from the vehicle. The agency did not separately consider the continuous belt option, but treated it together with the ignition interlock device in a category it titled "Option of Adopting Use-Compelling Features." 46 Fed. Reg. 53424  (1981). The agency was concerned that use-compelling devices would "complicate the extrication of [an] occupant from his or her car." Ibid. "[T]o require that passive belts contain use-compelling features," the agency observed, "could be counterproductive [, given] . . . widespread, latent and irrational fear in many members of the public that they could be trapped by the seat belt after a crash." Ibid. In addition, based on the experience with the ignition interlock, the agency feared that use-compelling features might trigger adverse public reaction.56
By failing to analyze the continuous seatbelts option in its own right, the agency has failed to offer the rational connection between facts and judgment required to pass muster under the arbitrary-and-capricious standard. We agree with the Court of Appeals that NHTSA did not suggest that the emergency release mechanisms used in nondetachable belts are any less effective for emergency egress than the buckle release system used in detachable belts. In 1978, when General Motors obtained the agency's approval to install a continuous passive belt, it assured the agency that nondetachable belts with spool releases were as safe as detachable belts with buckle releases. 43 Fed. Reg. 21912, 21913-21914 (1978). NHTSA was satisfied that this belt design assured easy extricability: "[t]he agency does not believe that the use of [such] release mechanisms will cause serious occupant egress problems . . . ." Id., at 52493, 52494. While the agency is entitled to change its view on the acceptability of continuous passive belts, it is obligated to explain its reasons for doing so.57
The agency also failed to offer any explanation why a continuous passive belt would engender the same adverse public reaction as the ignition interlock, and, as the Court of Appeals concluded, "every indication in the record points the other way." 220 U. S. App. D. C., at 198, 680 F. 2d, at 234.  We see no basis for equating the two devices: the continuous belt, unlike the ignition interlock, does not interfere with the operation of the vehicle. More importantly, it is the agency's responsibility, not this Court's, to explain its decision.58
"An agency's view of what is in the public interest may change, either with or without a change in circumstances. But an agency changing its course must supply a reasoned analysis . . . ." Greater Boston Television Corp. v. FCC, 143 U. S. App. D. C. 383, 394, 444 F. 2d 841, 852 (1970) (footnote omitted), cert. denied, 403 U. S. 923 (1971). We do not accept all of the reasoning of the Court of Appeals but we do conclude that the agency has failed to supply the requisite "reasoned analysis" in this case. Accordingly, we vacate the judgment of the Court of Appeals and remand the cases to that court with directions to remand the matter to the NHTSA for further consideration consistent with this opinion.60
I join Parts I, II, III, IV, and V-A of the Court's opinion. In particular, I agree that, since the airbag and continuous  spool automatic seatbelt were explicitly approved in the Standard the agency was rescinding, the agency should explain why it declined to leave those requirements intact. In this case, the agency gave no explanation at all. Of course, if the agency can provide a rational explanation, it may adhere to its decision to rescind the entire Standard.63
I do not believe, however, that NHTSA's view of detachable automatic seatbelts was arbitrary and capricious. The agency adequately explained its decision to rescind the Standard insofar as it was satisfied by detachable belts.64
The statute that requires the Secretary of Transportation to issue motor vehicle safety standards also requires that "[e]ach such . . . standard shall be practicable [and] shall meet the need for motor vehicle safety." 15 U. S. C. § 1392(a) (1976 ed., Supp. V). The Court rejects the agency's explanation for its conclusion that there is substantial uncertainty whether requiring installation of detachable automatic belts would substantially increase seatbelt usage. The agency chose not to rely on a study showing a substantial increase in seatbelt usage in cars equipped with automatic seatbelts and an ignition interlock to prevent the car from being operated when the belts were not in place and which were voluntarily purchased with this equipment by consumers. See ante, at 53, n. 16. It is reasonable for the agency to decide that this study does not support any conclusion concerning the effect of automatic seatbelts that are installed in all cars whether the consumer wants them or not and are not linked to an ignition interlock system.65
The Court rejects this explanation because "there would seem to be grounds to believe that seatbelt use by occasional users will be substantially increased by the detachable passive belts," ante, at 54, and the agency did not adequately explain its rejection of these grounds. It seems to me that the agency's explanation, while by no means a model, is adequate. The agency acknowledged that there would probably be some increase in belt usage, but concluded that the increase would be small and not worth the cost of mandatory  detachable automatic belts. 46 Fed. Reg. 53421-53423 (1981). The agency's obligation is to articulate a " `rational connection between the facts found and the choice made.' " Ante, at 42, 52, quoting Burlington Truck Lines, Inc. v. United States, 371 U. S. 156, 168 (1962). I believe it has met this standard.66
The agency explicitly stated that it will increase its educational efforts in an attempt to promote public understanding, acceptance, and use of passenger restraint systems. 46 Fed. Reg. 53425 (1981). It also stated that it will "initiate efforts with automobile manufacturers to ensure that the public will have [automatic crash protection] technology available. If this does not succeed, the agency will consider regulatory action to assure that the last decade's enormous advances in crash protection technology will not be lost." Id., at 53426.67
The agency's changed view of the standard seems to be related to the election of a new President of a different political party. It is readily apparent that the responsible members of one administration may consider public resistance and uncertainties to be more important than do their counterparts in a previous administration. A change in administration brought about by the people casting their votes is a perfectly reasonable basis for an executive agency's reappraisal of the costs and benefits of its programs and regulations. As long as the agency remains within the bounds established by Congress, it is entitled to assess administrative records and evaluate priorities in light of the philosophy of the administration.68
 Together with No. 82-355, Consumer Alert et al. v. State Farm Mutual Automobile Insurance Co. et al.; and No. 82-398, United States Department of Transportation et al. v. State Farm Mutual Automobile Insurance Co. et al., also on certiorari to the same court.69
 Briefs of amici curiae urging affirmance were filed by Dennis J. Barbour for the American College of Preventive Medicine et al.; by Nathan Lewin for the American Insurance Association; by Philip R. Collins and Thomas C. McGrath, Jr., for the Automotive Occupant Protection Association; by Alexandra K. Finucane for the Epilepsy Foundation of America et al.; by Katherine I. Hall for the Center for Auto Safety et al.; by Simon Lazarus III for Mothers Against Drunk Drivers; and by John H. Quinn, Jr., and John Hardin Young for the National Association of Insurance Commissioners.70
 National Safety Council, 1982 Motor Vehicle Deaths By States (May 16, 1983).71
 The Senate Committee on Commerce reported:72
"The promotion of motor vehicle safety through voluntary standards has largely failed. The unconditional imposition of mandatory standards at the earliest practicable date is the only course commensurate with the highway death and injury toll." S. Rep. No. 1301, 89th Cong., 2d Sess., 4 (1966).
 The Secretary's general authority to promulgate safety standards under the Act has been delegated to the Administrator of the National Highway Traffic Safety Administration (NHTSA). 49 CFR § 1.50(a) (1982). This opinion will use the terms NHTSA and agency interchangeably when referring to the National Highway Traffic Safety Administration, the Department of Transportation, and the Secretary of Transportation.74
 Early in the process, it was assumed that passive occupant protection meant the installation of inflatable airbag restraint systems. See 34 Fed. Reg. 11148 (1969). In 1971, however, the agency observed that "[s]ome belt-based concepts have been advanced that appear to be capable of meeting the complete passive protection options," leading it to add a new section to the proposed standard "[t]o deal expressly with passive belts." 36 Fed. Reg. 12859.75
 The court did hold that the testing procedures required of passive belts did not satisfy the Act's requirement that standards be "objective." 472 F. 2d, at 675.76
 Because such a passive restraint standard was not technically in effect at this time due to the Sixth Circuit's invalidation of the testing requirements, see n. 5, supra, the issue was not submitted to Congress until a passive restraint requirement was reimposed by Secretary Adams in 1977. To comply with the Amendments, NHTSA proposed new warning systems to replace the prohibited continuous buzzers. 39 Fed. Reg. 42692 (1974). More significantly, NHTSA was forced to rethink an earlier decision which contemplated use of the interlocks in tandem with detachable belts. See n. 13, infra.77
 No action was taken by the full House of Representatives. The Senate Committee with jurisdiction over NHTSA affirmatively endorsed the Standard, S. Rep. No. 95-481 (1977), and a resolution of disapproval was tabled by the Senate. 123 Cong. Rec. 33332 (1977).78
 Judge Edwards did not join the majority's reasoning on these points.79
 The Department of Transportation suggests that the arbitrary-and-capricious standard requires no more than the minimum rationality a statute must bear in order to withstand analysis under the Due Process Clause. We do not view as equivalent the presumption of constitutionality afforded legislation drafted by Congress and the presumption of regularity afforded an agency in fulfilling its statutory mandate.80
 For example, an overwhelming majority of the Members of the House of Representatives voted in favor of a proposal to bar NHTSA from spending funds to administer an occupant restraint standard unless the standard permitted the purchaser of the vehicle to select manual rather than passive restraints. 125 Cong. Rec. 36926 (1979).81
 While NHTSA's 1970 passive restraint requirement permitted compliance by means other than the airbag, 35 Fed. Reg. 16927, "[t]his rule was a de facto air bag mandate since no other technologies were available to comply with the standard." Graham & Gorham, NHTSA and Passive Restraints: A Case of Arbitrary and Capricious Deregulation, 35 Ad. L. Rev. 193, 197 (1983). See n. 4, supra.82
 Although the agency suggested that passive restraint systems contain an emergency release mechanism to allow easy extrication of passengers in the event of an accident, the agency cautioned that "[i]n the case of passive safety belts, it would be required that the release not cause belt separation, and that the system be self-restoring after operation of the release." 36 Fed. Reg. 12866 (1971).83
 In April 1974, NHTSA adopted the suggestion of an automobile manufacturer that emergency release of passive belts be accomplished by a conventional latch — provided the restraint system was guarded by an ignition interlock and warning buzzer to encourage reattachment of the passive belt. 39 Fed. Reg. 14593. When the 1974 Amendments prohibited these devices, the agency simply eliminated the interlock and buzzer requirements, but continued to allow compliance by a detachable passive belt.84
 See, e. g., Comments of Chrysler Corp., Docket No. 69-07, Notice 11 (Aug. 5, 1971) (App. 2491); Chrysler Corp. Memorandum on Proposed Alternative Changes to FMVSS 208, Docket No. 44, Notice 76-8 (1976) (App. 2241); General Motors Corp. Response to the Dept. of Transportation Proposal on Occupant Crash Protection, Docket No. 74-14, Notice 08 (May 27, 1977) (App. 1745). See also Chrysler Corp. v. Department of Transportation, 472 F. 2d 659 (CA6 1972).85
 The Department of Transportation expresses concern that adoption of an airbags-only requirement would have required a new notice of proposed rulemaking. Even if this were so, and we need not decide the question, it would not constitute sufficient cause to rescind the passive restraint requirement. The Department also asserts that it was reasonable to withdraw the requirement as written to avoid forcing manufacturers to spend resources to comply with an ineffective safety initiative. We think that it would have been permissible for the agency to temporarily suspend the passive restraint requirement or to delay its implementation date while an airbag mandate was studied. But, as we explain in text, that option had to be considered before the passive restraint requirement could be revoked.86
 Between 1975 and 1980, Volkswagen sold approximately 350,000 Rabbits equipped with detachable passive seatbelts that were guarded by an ignition interlock. General Motors sold 8,000 1978 and 1979 Chevettes with a similar system, but eliminated the ignition interlock on the 13,000 Chevettes sold in 1980. NHTSA found that belt usage in the Rabbits averaged 34% for manual belts and 84% for passive belts. RIA, at IV-52, App. 108. For the 1978-1979 Chevettes, NHTSA calculated 34% usage for manual belts and 72% for passive belts. On 1980 Chevettes, the agency found these figures to be 31% for manual belts and 70% for passive belts. Ibid.87
 "NHTSA believes that the usage of automatic belts in Rabbits and Chevettes would have been substantially lower if the automatic belts in those cars were not equipped with a use-inducing device inhibiting detachment." Notice 25, 46 Fed. Reg. 53422 (1981).88
 NHTSA commissioned a number of surveys of public attitudes in an effort to better understand why people were not using manual belts and to determine how they would react to passive restraints. The surveys reveal that while 20% to 40% of the public is opposed to wearing manual belts, the larger proportion of the population does not wear belts because they forgot or found manual belts inconvenient or bothersome. RIA, at IV-25, App. 81. In another survey, 38% of the surveyed group responded that they would welcome automatic belts, and 25% would "tolerate" them. See RIA, at IV-37, App. 93. NHTSA did not comment upon these attitude surveys in its explanation accompanying the rescission of the passive restraint requirement.89
 Four surveys of manual belt usage were conducted for NHTSA between 1978 and 1980, leading the agency to report that 40% to 50% of the people use their belts at least some of the time. RIA, at IV-25, App. 81.90
 The Court of Appeals noted previous agency statements distinguishing interlocks from passive restraints. 42 Fed. Reg. 34290 (1977); 36 Fed. Reg. 8296 (1971); RIA, at II-4, App. 30.91
 Petitioners construe the Court of Appeals' order of August 4, 1982, as setting an implementation date for Standard 208, in violation of Vermont Yankee's injunction against imposing such time constraints. Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, Inc., 435 U. S. 519, 544-545 (1978). Respondents maintain that the Court of Appeals simply stayed the effective date of Standard 208, which, not having been validly rescinded, would have required mandatory passive restraints for new cars after September 1, 1982. We need not choose between these views because the agency had sufficient justification to suspend, although not to rescind, Standard 208, pending the further consideration required by the Court of Appeals, and now, by us.92
 Of course, a new administration may not refuse to enforce laws of which it does not approve, or to ignore statutory standards in carrying out its regulatory functions. But in this case, as the Court correctly concludes, ante, at 44-46, Congress has not required the agency to require passive restraints.