Opinion for the Court filed by Circuit Judge WALD.
Dissenting opinion filed by Senior Circuit Judge BAZELON in which Circuit Judges TAMM and WILKEY join.
*1563WALD, Circuit Judge.In this appeal, the General Motors Corporation (GM) challenges the authority of the Environmental Protection Agency (EPA or Agency) to require manufacturers, as a part of a recall pursuant to the Clean Air Act, to repair cars and engines, which are members of the recall class, but which have exceeded their statutory “useful lives” of five years or 50,000 miles at the time of repair. In 1980, the EPA promulgated what it termed an “interpretive rule,” embodying its consistent regulatory practice since the enactment of the Clean Air Act Amendments of 1970, requiring manufacturers to repair all members of a recall class, regardless of their age or mileage at the time of repair. Soon thereafter, in accordance with the interpretative rule, the agency ordered GM to submit a remedial plan for a class of recalled Cadillacs that included all class members regardless of age or mileage. In its petitions to this court, GM contends that (1) the EPA rule is a legislative, not interpretative, rule, and therefore the rule is void for failure to comply with the notice and comment procedures set down by the Administrative Procedure Act, and (2) in any event, the rule runs counter to the recall provision of the Clean Air Act and therefore is an invalid interpretation of the statute. GM consequently challenges the validity of both the interpretative rule and the agency order that applied the rule to the recall of GM automobiles. For the reasons stated below, we hold that the EPA rule is a valid interpretative rule, supported by the language, purpose and legislative history of section 207(c) of the Clean Air Act Amendments of 1970. We accordingly deny GM’s petitions.
I. Background
On March 21, 1977, after an EPA investigation and testing by both EPA and GM, the Administrator of the EPA notified GM pursuant to section 207(c)(1) of the Clean Air Act that the class of 1975 Cadillacs with type 230-carburetors did not conform with federal standards for carbon monoxide emission, and ordered GM to “submit a plan for recalling and remedying the nonconformity of the vehicles.”1 At the same time, the Administrator expressed concern that another set of 1975 Cadillacs — with type 193-carburetors — also exhibited nonconformity with the carbon monoxide standards, but he refrained from initiating a formal recall of that class in order to give GM an opportunity to substantiate its “strong protestations” that this latter class did not violate the standards. See Letter from Douglas M. Costle, EPA Administrator, to E.M. Estes, President of GM (March 21, 1977), reprinted in Joint Appendix (J.A.) at 122-24.
Five weeks later, GM filed a proposed remedial plan for the 230-earburetor Cadillacs, and as a part of the plan offered to recall the 193-carburetor Cadillacs voluntarily.2 See J.A. at 126-27. EPA did not approve the plan, however, and negotiations continued for two years between the agency and GM over the effectiveness of the proposed repairs and the manufacturer’s obligation to audit the performance of *1564completed repairs. Finally, on December 26, 1979, the EPA withdrew its demand that GM conduct an audit, deciding instead to conduct the audit itself, and approved GM’s amended remedial plan.3
At that time, however, GM urged the EPA to cancel the recall, arguing that “as a direct result of the delays incurred since our first remedial plan was submitted, a point of drastically diminished returns has been reached.” Letter from T.M. Fisher, Automotive Emission Control Director for GM, to Benjamin R. Jackson, EPA Deputy Assistant Administrator for Mobile Source and Noise Enforcement (Feb. 5, 1980), reprinted in J.A. at 258. More specifically, GM contended for the first time that “only those vehicles within the lesser of five years or 50,000 miles of operation at the time of presentation to the dealer for repairs will receive the [remedial repairs] at General Motors [sic] expense.”4 Id., reprinted in J.A. at 259. GM thus believed that so few Cadillacs in the class would be subject to recall because of their age or mileage that the recall should not be implemented at all.
On May 30, 1980, the EPA promulgated the rule at the center of dispute in this case. See 45 Fed.Reg. 36,396 (May 30, 1980). Under the rule, all remedial plans filed under section 207(c) of the Clean Air Act “shall provide that the manufacturer will remedy, • at the manufacturer’s expense, all properly maintained and used vehicles which experienced the nonconformity during their useful lives regardless of their age or mileage at the time of repair.” 40 C.F.R. Subpart S App. A (emphasis added). Applying this rule to the recall of the 1975 Cadillacs, the EPA on June 23, 1980, approved GM’s Cadillac remedial plan insofar as it applied to vehicles within their useful lives at the time of repair, and ordered GM to submit a plan for the repair of automobiles within the recall class “which failed or will fail to conform to applicable emission standards during their useful lives but will be beyond their useful lives at the time of repair.” Letter from Charles N. Freed, EPA Acting Assistant Deputy Administrator for Mobile Source, Noise and Radiation Enforcement, to T.M. Fisher, Director of Automotive Emission Control for GM (June 23, 1980), reprinted in J.A. at 263.
GM then petitioned this court, challenging both the May 30 rule and the June 30 recall determinations.5 GM argues that (1) the May 30 rule constituted a legislative rulemaking, despite EPA’s characterization of the rule as “interpretive,” and therefore the EPA failed to follow the notice and comment procedures required before such rules may be promulgated, and (2) by requiring the repair of automobiles beyond their useful lives, the rule exceeds the EPA’s authority under section 207(c) of the Clean Air Act. For the reasons explained below, we disagree with both of GM’s contentions.
II. The EPA Rule Is an Interpretative Rule
EPA styled its May 30 rule as an “interpretive rule.” See 45 Fed.Reg. 36,396 (May 30, 1980). As a preliminary matter, GM challenges this characterization, contending that the May 30 rule is, on the *1565contrary, a “legislative rule” and is consequently void for failure to follow the notice and comment procedures required for the promulgation of such rules. See 5 U.S.C. § 553. Because we find that EPA properly categorized the May 30 rule, we conclude that notice and comment procedures were not required. See id. § 553(b)(A).6
In a turn of phrase particularly apt in this case, the distinction between legislative and nonlegislative rules has been described as “enshrouded in considerable smog.” American Bus Association v. ICC, 627 F.2d 525, 529 (D.C.Cir.1980) (quoting Noel v. Chapman, 508 F.2d 1023, 1030 (2d Cir.1975) (discussing definition of “general statement of policy”)). Nonetheless, there are certain general principles that aid reviewing courts in making the determination whether a given rule is legislative or interpretative. First, the agency’s own label, while relevant, is not dispositive. See, e.g., Chamber of Commerce v. Occupational Safety and Health Administration, 636 F.2d 464, 468 (D.C.Cir.1980); Citizens to Save Spencer County v. EPA, 600 F.2d 844, 879 n. 171 (D.C.Cir.1979). An interpretative rule simply states what the administrative agency thinks the statute means, and only “ ‘reminds’ affected parties of existing duties.” Citizens to Save Spencer County, 600 F.2d at 876 & n. 153; ste also Chamber of Commerce, 636 F.2d at 469. On the other hand, if by its action the agency intends to create new law, rights or duties, the rule is properly considered to be a legislative rule. See, e.g., American Postal Workers Union v. United States Postal Service, 707 F.2d 548, 558-59 (D.C.Cir.1983); Citizens to Save Spencer County, 600 F.2d at 876.
In light of these general principles, we find that the May 30 rule constitutes an interpretative rule. We note, to begin with, that the agency regarded its rule as interpretative. See 45 Fed.Reg. 36,396 (May 30, 1980) (“Action: Interpretive Rule”). Moreover, EPA’s entire justification for the rule is comprised of reasoned statutory interpretation, with reference to the language, purpose and legislative history of section 207(c). See id. at 36,397-98. Indeed, the language of the rule itself indicates its interpretative nature. See 40 C.F.R. Subpart S App. A (“The purpose of this rule is to set forth EPA’s interpretation ... under section 207(c)(1) of the Clean Air Act....”) Finally, and most importantly, the rule did not create any new rights or duties; instead, it simply restated the consistent practice of the agency in conducting recalls pursuant to section 207(c).7
Accordingly, we hold that the May 30 rule was an interpretative rule. The *1566EPA therefore was not required to follow notice and comment procedures prior to the adoption of the rule. See 5 U.S.C. § 553(b)(A). We now consider whether the interpretative rule is valid under section 207(c) of the Clean Air Act.
III. The EPA Rule Is a Permissible Interpretation op Section 207
The EPA’s May 30 rule finds substantial support, and no contradiction, in the language and legislative history of section 207 of the Clean Air Act. Moreover, the May 30 rule effectuates important public policy goals embodied in the Act. Accordingly, we uphold the validity of the EPA rule at issue in this case.
The Supreme Court has recently outlined our proper task in reviewing an administrative construction of a statute that the agency administers. First, we must determine whether Congress “has directly spoken to the precise question at issue.” Chevron, U.S.A. v. National Resources Defense Council, — U.S. — at —, 104 S.Ct. 2778 at 2781, 81 L.Ed.2d 694 (1984). If the administrative construction *1567runs counter to clear congressional intent, then the reviewing court must reject it. See id. at — n. 9, 104 S.Ct. at 2782 n. 9; see also FEC v. Democratic Senatorial Campaign Committee, 454 U.S. 27, 32, 102 S.Ct. 38, 42, 70 L.Ed.2d 23 (1981). On the other hand, if the administrative construction does not contravene clearly discernible legislative intent, then the reviewing court “does not simply impose its own construction on the statute.” Chevron, slip op. — U.S. at —, 104 S.Ct. at 2782. Instead, we then must conduct the “narrower inquiry into whether the [agency’s] construction was ‘sufficiently reasonable’ to be accepted by a reviewing court.”8 Democratic Senatorial Campaign Committee, 454 U.S. at 39, 102 S.Ct. at 46; see Chevron, slip op. — U.S. at —, 104 S.Ct. at 2782.
We begin our review with an examination of the statute. Section 207(c)(1) provides, in pertinent part:
If the Administrator determines that a substantial number of any class or category of vehicles or engines, although properly maintained and used, do not conform to the [EPA emission standards], when in actual use throughout their useful life (as determined under section 7521(d) of this title), he shall immediately notify the manufacturer thereof of such nonconformity, and he shall require the manufacturer to submit a plan for remedying the nonconformity of the vehicles or engines with respect to which such notification is given.
42 U.S.C. § 7541(c)(1). The EPA interprets this provision to authorize the recall of all members of a nonconforming class, except those not “properly maintained and used,” regardless of the age or mileage of any given member. We think the statute provides ample support for such an interpretation.
The statute requires manufacturers to submit remedial plans for “vehicles or engines with respect to which ... notification is given,” and directs the EPA Administrator to give such notification with respect to a “class or category of vehicles or engines,” a substantial number of which exhibited nonconformity during their useful lives. The statute thus provides for notice on a classwide basis and further requires the manufacturer to remedy all cars with respect to which such notice is given. In light of this class-based orientation of section 207(c)(1), we think the EPA reasonably mandated that remedial plans be designed to cover all members of a recall class. The agency reasonably required the manufacturers to include a ear or engine in their remedial plans if (1) a substantial number of cars within a given class exhibited nonconformity during their useful lives, and (2) the car is a member of that class. Section 207(c) affords an ample basis for this requirement.
GM nevertheless argues that individual vehicles and engines that have exceeded their useful lives cannot possibly exhibit a “nonconformity,” because the Clean Air *1568Act’s emission standards apply only to vehicles and engines within their useful lives. See 42 U.S.C. § 7521(a)(1). Since section 207(c) requires manufacturers to devise remedial plans so that a recall will “remedy the nonconformity” of the class, GM concludes that its remedial obligations extend only to those members of a recall class that are within their useful lives at the time of repair.
However, while it is true that the emission standards apply only to individual cars and engines during their useful lives, it does not follow that only those cars and engines within their useful lives may be recalled. Indeed, the question presented by this case is whether a member vehicle or engine of a recall class that has exceeded its useful life should nevertheless be repaired under section 207(c) at the manufacturer’s expense. That question is not settled by the definition of “useful life”; rather, it turns on the intended scope and purpose of the recall provision. As we have explained above, section 207(c) provides for classwide remedies of classwide defects. Accordingly, EPA reasonably reads the section to require the manufacturers to submit plans to “remedy the nonconformity” of the recall class. After all, a defective class — the members of which do not satisfy the emission standards during their useful lives — can be considered “nonconforming.” Hence the May 30 rule is not precluded by the statute’s definition of “useful life.”
Moreover, assuming arguendo that only individual cars and vehicles within their useful lives could exhibit a “nonconformity,” a plan “for remedying the nonconformity” could still include cars or engines beyond their useful lives at the time of repair. Through the recall scheme, Congress obviously intended the EPA to conduct tests on representative samplings of cars and engines, and to base its decision to recall on such tests.9 Unless Congress wanted the EPA to test every car in a class — an absurd prospect — it was willing to allow a statistical inference from representative testing that all members of the class exhibit a nonconformity during their useful lives. Once it is established that a car was out of compliance with emission standards during its useful life, a reasonable remedy could include the repair of the car to compensate for the pollution caused during the time of its violation. Therefore, once the EPA establishes that members of a recall class were generally nonconforming during their useful lives, a reasonable method of “remedying the nonconformity” includes the repair of all members of the class, even if some had exceeded their useful lives.10 We *1569note that, under the EPA rule, if a certain car or subclass of cars were in compliance during the five year or 50,000 mile period, then the manufacturer need not repair that car or subclass of cars.11 See 45 Fed.Reg. 36,397 n. 2 (May 30, 1980) (“the manufacturer is not responsible to remedy the nonconformity of a vehicle which, although part of the recall class, experienced the nonconformity only after expiration of the vehicle’s useful life”).
The legislative history similarly supports EPA’s interpretative rule. The Report of the Senate Public Works Committee — upon which GM heavily relies, see infra at p. 1570 — refers to the recall of an entire “model or class” of vehicles or engines. S.Rep. No. 1196, 91st Cong., 2d Sess. 29 (1970). The Senate Committee bill designed two methods for EPA to assure in-use compliance with emission standards. First, after the development of a “quick test” method, the EPA could test the continuing compliance of “individual vehicles on the road.” Id. Alternatively, the EPA could conduct more intensive examinations of “representative samplefs] of a model or class,” and, after a finding of classwide noncompliance, EPA “could require the manufacturer to recall that model or class.” Id.; see id. at 111 (section 207 of S. 4358) (if EPA finds that “statistically representative samples of any class or category of vehicles or vehicle engines ... do not conform,” then all “vehicle engines included within the class or category” should be notified); id. at 62-63 (section-by-seetion analysis) after the EPA “discovers defects through testing” of a class of vehicles or engines, it “shall order the manufacturer to notify ... purchasers of the defect ”) (emphasis added).
Furthermore, when Congress revisited section 207(c) in 1977, the discussion assumed that the recall provision extended to all member vehicles or engines of the nonconforming classes. See, e.g., H.Rep. No. 294, 95th Cong., 1st Sess. 497, U.S.Code Cong. & Admin.News 1977, pp. 1077, 1459 (“if a substantial number of systems fail during their on-the-road operation, the EPA can recall the entire lot for repair at the manufacturer’s expense”); id. at 498, U.S. Code Cong. & Admin.News 1977, p. 1460 (noting the possibility of “a recall of all of that model vehicle or engine type”); 6 Environmental Policy Division of the Congressional Research Service of the Library for the Comm, on Environment and Public Works, 95th Cong., 2d Sess. 4540 (1978) (preliminary statement of Sen. Bentsen) (“section 207(c) authorizes the Agency to require the automaker to recall a ' given model run for needed repairs if the Agency determines that a substantial number of that model or engine type do not conform to the standards when in actual use”) (emphasis added); id. at 1220 (preliminary statement of Sen. Riegle) (“If a substantial number of systems fail during their on-the-road operation, the EPA can recall the entire lot for repair at the manufactur*1570er’s expense”). Given these indicia of the congressional understanding of section 207(c), and the contemporaneous administrative policy (as evidenced by manufacturers' remedial plans, see supra at p. 1563), it is reasonable to assume that Congress reaffirmed the EPA’s understanding of its recall authority when it revisited and ratified the recall provision in 1977. There is not a shred of evidence to the contrary.12
Nonetheless, GM contends that one bit of legislative history contradicts the EPA’s interpretation. Specifically, GM heavily relies upon one passage from the Report of the Senate Public Works Committee. In discussing a provision that was the precursor to section 207 as finally enacted, the Report said: S.Rep. No. 1196, supra, at 31. GM believes that this passage constitutes a clear indication that Congress intended to limit the manufacturers’ recall obligations to exclude cars and engines beyond their useful lives.
The Committee also recognizes the difficulty in any recall provision of notifying the owners of vehicles. The Committee expects that the manufacturer would not only depend on the files of the franchise dealer, but would, to the extent practicable, use State motor vehicle department registration files to obtain the names and addresses of subsequent purchasers of cars. By establishing a 50,000 mile, no year lifetime for the purpose of warranty, the Committee did not intend to relieve the automobile manufacturers of their responsibility to notify owners of older cars. The 50,000-mile period can be assumed to be 4 to 5 years and the manufacturer should be expected to notify any owner of a vehicle that is five years old or less as to failure to continue to perform to the standard. A decision not to require the manufacturer to repair the vehicle could be made after notice and after finding that the vehicle had exceeded the 50,000-mile warranty period.
We disagree. In the first place, we would be very hesitant to overrule a reasonable agency interpretation, which enjoys support in the statutory language and other portions of legislative history, on the basis of a single, and rather ambiguous, passage from a committee report.13 Furthermore, as explained below, we find that this passage in no way undermines the EPA’s interpretation of section 207, because the Senate bill under consideration at the time the Committee wrote its report did not contain a recall provision similar to the provision finally embodied in section 207. Accordingly, we conclude that nothing in the legislative history runs counter to the EPA’s May 30 rule.
The Senate bill did not impose upon manufacturers any recall repair obligation; it provided only for the warranty repair obligation. Section 207(d)(2) of the bill merely directed the EPA to “order the manufacturer to provide prompt notification ... [to] purchasers of all ... vehicles or vehicle engines included in the class or category” that the EPA found to be in noncompliance with emission standards. See S.Rep. 1196, *1571supra, at 111. Once the purchaser received such a notice, he could have sought repair at the manufacturer’s expense only under the warranty mandated by section 207(c) of the bill. See id. at 110. As in the final Act, the bill expressly limited the life of the warranty to the useful life of the vehicle or engine. See id. (section 207(c) of S. 4358); see also 42 U.S.C. § 7541(a) (“the manufacturer ... shall warrant ... that such vehicle or engine is ... free from defects in materials and workmanship which cause such vehicle or engine to fail to conform with applicable regulations for its useful life ”) (emphasis supplied). Thus, the Senate bill’s so-called “recall” provision merely required manufacturers to notify purchasers of a class of vehicles of a defect, and relied on the warranty repair obligation — which was expressly limited in duration to the useful life of the vehicle or engine — to require the manufacturers to pay for repairs. Under this scheme, it might well have been reasonable to limit the “recall” repair obligation to include only cars and engines within their useful lives.
However, the law as finally enacted, in contrast to the Senate bill, creates an independent recall repair obligation in addition to the warranty obligation. Section 7541(c) of title 42 requires the manufacturer of a recalled class to submit and comply with a remedial plan for that class. Unlike the warranty repair obligation, which covers only failure to conform with emission standards “for [the individual vehicle or engine’s] useful life,” 42 U.S.C. § 7541(a), the recall repair obligation contains no such limitation, and can reasonably be considered to extend to all vehicles or engines in the recall class. Thus, the passage from the Senate report explains an earlier version of section 207, which is irrelevant to the final version of the Act for the purposes of determining the scope of the recall repair obligation. We decline to overrule an otherwise reasonable agency interpretation on such an attenuated basis.
The EPA’s rule also furthers the purposes of the Congress. The Clean Air Act’s unequivocal directive is “to protect and enhance the quality of the Nation’s air resources so as to promote the public health and welfare and the productive capacity of its population.” 42 U.S.C. § 7401(b)(1). The May 30 rule obviously effectuates this purpose. When the EPA discovers a defect that results in excessive pollution in a class of cars or engines, the rule assures that the manufacturer must fashion a remedy to fit the violation: the remedial plan must include “all properly used and maintained nonconforming vehicles in the subject class regardless of their age or mileage at the time of repair.” 45 Fed.Reg. 36,397 (May 30, 1980).
This interpretation makes good practical sense because, no matter what their age at the time of actual recall, nonconforming cars have been riding the roads in violation of pollution standards. Before the EPA institutes a recall, it must make a determination that the members of the recall class have generally exceeded emission standards during their useful lives. See 42 U.S.C. § 7541(c)(1). The agency therefore may reasonably require manufacturers to include older cars and engines in their plans to “remedy the nonconformity” of the class. Cars and engines that failed to conform with emission standards during their useful lives do not cease to be a pollution hazard thereafter; in fact, it is sensible to presume that they are more likely to be a hazard afterward than conforming cars and engines.14 To the degree that the members of a nonconforming class can be repaired to decrease their pollution potential even after their useful lives have *1572expired, the public is benefitted. By the same token, to the degree that they elude correction both during and beyond their useful lives, the public is cheated.15 We therefore do not believe that older members of a recall class must be granted an exemption from plans for the remedy of a classwide defect simply because of their age or mileage at the time the recall is initiated or implemented, especially when, under such an interpretation, the number of cars and engines escaping repair could be increased as a result of manufacturer or administrative delays in getting the recall underway.16
IV. Conclusion
For the reasons stated above, we hold that the EPA’s May 30 rule is an interpretative rule. This rule enjoys adequate support in the language, purpose and legislative history of the Clean Air Act. GM’s petitions are therefore denied.
So ordered.
. Letter from Douglas M. Costle, EPA Administrator, to E.M. Estes, President of GM (March 21, 1977), reprinted in Joint Appendix (J.A.) at 122-24. While it was and is the agency's policy to encourage voluntary recalls, the Administrator "fe[lt] compelled to issue th[e] order” because "General Motors’ acquiescence to recall comes at the end of more than two years of EPA investigation and GM opposition, and after assemblage of the most convincing case yet for recall." Id. Moreover, the Administrator expressed “some concern over the likelihood of actual implementation by dealers of the remedy proposed by General Motors and wish[ed] to preserve [the] right to approve the remedy to be actually pursued.” Id.
. The remedial plan initially submitted by GM covered "all 1975 Cadillac models except those built with California Emission Controls.” Remedial Plan for 1975 Cadillacs (April 27, 1977), reprinted in J.A. at 129. The plan specifically provided that "[a] vehicle affected will be eligible for repair under the remedial plan, regardless of vehicle age, mileage, or owner." Id., reprinted in J.A. at 131.
. See Letter from Benjamin R. Jackson, Deputy Assistant Administrator for Mobile Source and Noise Enforcement, to T.M. Fisher, Director of GM Automotive Emission Control (Dec. 26, 1979), reprinted in J.A. at 252-55.
. GM contends that its 1980 letter merely restated a position it had taken a year earlier in negotiations with the EPA. See GM Brief at 5. EPA, however, asserts that the letter represents "the first time that GM had proposed to impose a 5/50 [useful life] limitation on a remedial plan.” EPA Brief at 11. Nothing in the record shows that GM presented this limitation prior to its 1980 letter.
. On November 26, 1980, the EPA published a notice in the Federal Register declaring its partial approval and partial disapproval of GM’s remedial plan to be "final.” See 45 Fed.Reg. 78798 (Nov. 26, 1980). GM subsequently filed a protective petition, consolidated with its previous appeals, seeking review of this Federal Register notice.
. The APA exempts from its notice and comment procedures “interpretative rules, general statements of policy, or rules of agency organization, procedure, or practice.” 5 U.S.C. § 553(b)(A). Similarly, the Clean Air Act exempts from its special rulemaking procedures "any rule or circumstance, referred to in sub-paragraphs (A) or (B) of subsection 553(b) of title 5.” 42 U.S.C. § 7607(d)(1). Accordingly, interpretative rules are not subject to the procedural requirements set down by either the APA or the Clean Air Act.
. The dissent contests this proposition and calls it "a proposition upon which the majority places substantial weight.” Diss.Op. at 1574. At the outset, it should be made clear that this proposition is significant primarily in support of the conclusion that the EPA rule is interpretative, and not legislative, a proposition with which the dissent agrees. See Diss.Op. at 1573 & n. 3.
In its attempt to make light of the EPA’s historical adherence to its current interpretation, the dissent asserts that the EPA interpretation has only “a lifespan barely exceeding one year” because the sole controversy over the interpretation occurred in 1979. See Diss.Op. at 1574 n. 11. However, as the EPA observed when it issued its interpretative rule ”[i]n the past, manufacturers have not conditioned a vehicle’s eligibility for recall repair ... on the basis of the vehicle's age or mileage.” 45 Fed. Reg. at 36,397. Rather than diminishing the rule’s status as a longstanding EPA interpretation, the absence of any prior controversy over the repair of cars exceeding their useful lives demonstrates the reasonableness of the EPA’s interpretation of its own regulatory scheme. See, e.g., Esquire, Inc. v. Ringer, 591 F.2d 796, 801 (D.C.Cir.1978) (Bazelon, J.) (administrative interpretation deserves “controlling weight," particularly when it "has been consistently followed for a significant period of time”), cert. denied, 440 U.S. 908, 99 S.Ct. 1217, 59 L.Ed.2d 456 (1979); DeLano v. United States, 183 Ct.Cl. *1566379, 393 F.2d 517, 521 (1968) ("A strong indication of the reasonableness of that [administrative] construction is the fact that [the regulated parties] never challenged it” even though it was “consistently maintained by the agency.”). Apparently, virtually every manufacturer engaging in a recall has voluntarily included older cars in its own remedial plans.
The dissent mistakenly speculates that “manufacturers in previous recalls have failed to raise the question of liability for repair of cars beyond their useful lives because the involvement of such vehicles in the recalls has been de minimis." Diss.Op. at 1574 n. 12. The dissent here loses sight of the facts. Despite its view that "virtually all” of the recalls involved tiny percentages of older cars, in six out of the fifteen recalls described in detail in the record, see J.A. at 278-96, the EPA issued a recall notice to the manufacturer 34-55 months after the cars were on the market. Under EPA’s method of calculating the percentage of cars recalled that were beyond their useful lives, see J.A. at 275, these recalls involved large quantities of older cars. For example, the Pontiac recall, initiated in 1979, involved 1975 model year cars, 54% of which were beyond their useful lives. See J.A. at 293. Moreover, almost 30% of the 1976 Pontiacs involved in that recall had exceeded their useful lives at the time of manufacturer notice. Similarly, in 1978 EPA initiated a recall that included 1975 Fords, approximately 28% of which had exceeded their useful lives. See J.A. at 289. Also, in 1978 the EPA began a recall that included 1974 AMC cars sold in California, approximately 56% of which were beyond their useful lives. See J.A. at 291-92. Any manufacturer who "prophetically foresaw” the dissent’s position, Diss.Op. at 1574 n. 12, believing these recalls exceeded the legal scope of coverage, could reasonably have been expected to make an objection.
Moreover, at least three other recalls included substantial numbers of olders cars. See J.A. at 280 (1974 Pontiacs; approximately 22% beyond useful lives); J.A. at 288 (1975 Fords; approximately 16% beyond useful lives); J.A. at 296 (1977 Buicks; approximately 28.5% beyond useful lives).
The dissent obfuscates the issue by claiming that the above discussion ”neglect[s] the fact” that some of these recalls included younger cars in addition to older cars, and that the "AMC recall was complicated by the more stringent emissions standards” applicable only in California. Diss.Op. at 1574 n. 12. However interesting, these details are irrelevant to the dissent’s assertion that the manufacturers "failed to raise the question of liability for repair of cars beyond their useful lives because the involvement of such vehicles in the recalls has been de minimis." Id. The recall of younger cars in addition to older cars in no way detracts from the fact that there were substantial numbers of older cars involved in these recalls. The relevance of the California air emission standards to the dissent’s position is at best obscure.
In a final aside, the dissent invokes the fact that "owner responses to recalls drop dramatically with the age of the vehicle." Id. Such statistical sport, however, ignores the fact that the manufacturer incurs heavy costs — both financial and good will — simply by issuing its notice to owners. Moreover, while the response rate declines as time goes by, the percentage of older cars in the recall sales rises commensurately. See J.A. at 275, 298. Therefore, whatever the age of the recall class, the manufacturer would still have a substantial stake and ample incentive to raise the useful life issue.
In sum, the dissent fails in its attempts to transmogrify the absence of controversy over the EPA’s policy into the absence of any EPA policy on the useful life limitation. Taken individually or cumulatively, the dissent's efforts to invent post hoc rationalizations for the manufacturers’ failure to challenge the EPA’s policy are similarly unconvincing. The EPA has consistently approved only those remedial plans that included all cars and engines in the recall class, regardless of their age or mileage. It therefore constitutes a longstanding administrative policy.
. It is difficult to square the dissent’s position with these recent pronouncements by the Supreme Court. Although the dissent attempts to invoke the "plain meaning of the statute,” Diss. Op. at 1576, in truth it argues only that the useful life limitation should be imported into the recall provision. See id. at 1575-1576. However, nothing in the language or legislative history of the recall provision compels this result. See infra at pp. 1578-1580. The absence of a useful life limitation in the recall provision should be compared with the express limitation of the manufacturer’s liability under the warranty provision to emissions violations occurring during a vehicle’s "useful life.” Compare 42 U.S.C. § 7541(c)(1) (recall provision) with id. § 7541(a) (warranty provision). See infra at pp. 1570-1571.
Similarly, the dissent's extensive dicta, setting out what it dubs "some reasonable alternatives,” see Diss.Op. at 1582-1584, is particularly inappropriate in light of the Supreme Court’s recent warning that reviewing courts should respect the policy choices of administrative agencies. See Chevron, slip op. — U.S. at —, 104 S.Ct. at 2793-94. The question presented today is whether the agency’s action is a reasonable one in light of the existing statutory constraints, and it is not the job of the courts to propose rules that they would prefer the agency to adopt.
. The dissent erroneously concludes that the EPA rule "would expose the manufacturer to liability for repair virtually indefinitely." Diss.Op. at 1579. The statute, however, authorizes the Administrator to institute a recall only after determining that a substantial number of vehicles or engines in the class "do not conform to the [emission standards] when in actual use throughout their useful life." 42 U.S.C. § 7541(c)(1) (emphasis added). Since cars and engines older than five years of age are no longer within their useful lives, this statutory condition cannot be met "indefinitely." There is, at the very least, this definite time limitation on the Administrator: he must conclude his testing before the entire recall class has exceeded five years of age.
Moreover, special problems might arise when a proposed recall class includes numerous model years of a particular car or engine, some of which have entirely exceeded their useful lives at the time the EPA conducts its testing. See Diss.Op. at 1579 (hypothesizing such a class). The design of some features of a class of cars or engines might be altered from one model year to another. If these changes could affect the level of auto emissions, then the EPA’s testing of only the models within their useful lives might not be conclusive as to the nonconformity of the entire class proposed for recall. However, since the recall class presented in this case does not raise this problem, we need not decide this hypothetical. Suffice it to say that our reading of the statute amply permits the EPA to recall cars and engines regardless of their age at the time of recall. We do not address today the possible constraints upon the EPA in defining the reasonable scope of a recall class.
. At oral argument and in its supplemental brief, the EPA attested to its flexibility in devising appropriate and reasonable remedial schemes for older cars and engines. We do not find, and GM does not contend, that the remedy *1569proposed for the Cadillacs in this case is unreasonable.
. The dissent complains that “the EPA’s rule establishes an absolute and irrebutable presumption that all older cars were among the percentage failing to meet standards during their useful lives." Diss.Op. at 1582. However, it is the statute that establishes the presumption that the cars and engines within a class exhibiting substantial nonconformity are individually nonconforming. Moreover, this presumption is not "absolute and irrebuttable”; the manufacturer is free to come forward with evidence that an individual car or engine, or a particular subclass of cars and engines, were in fact conforming during their useful lives. The results of on-the-road testing would be particularly relevant to this inquiry.
This same proof problem, of course, would be present in the implementation of the dissent’s proposal to permit the Administrator to recall cars and engines that were within their useful lives at the time of the notice of nonconformity. Under this proposal, manufacturers would be required to repair cars and engines exceeding their useful lives without any showing that the particular cars or engines had violated the emission standards during their useful lives. The dissent’s proposal, to be practical, must also follow the statutory presumption that members of the recall class were nonconforming during their useful lives.
. In addition, EPA’s class-based interpretation of section 207 enjoys precedential support. This court, in one of the few judicial cases to construe the recall provision, observed that "[ujnlike the discovery and cure of nonconformity of individual vehicles under the warranty provisions, the remedy at this stage is recall of the entire class of vehicles in order to correct the design, material, or workmanship defect.” Chrysler Corp. v. EPA, 631 F.2d 865, 868 (D.C.Cir.) (emphasis in original), cert. denied, 449 U.S. 1021, 101 S.Ct. 589, 66 L.Ed.2d 483 (1980).
. As this court, sitting en banc, observed:
[I]t must be remembered that committee reports are not the law; they are only aids in interpreting statutory language and are useful only to the extent they fairly reflect congressional intent. Sometimes committee reports are not reliable guides to legislative intent, as, for example, where they contain statements that contradict the plain meaning of the statutory language or that conflict with the expressed purpose of the statute.
Jordan v. United States Dep’t of Justice, 591 F.2d 753, 767 (D.C.Cir.1978) (en banc) (footnotes omitted).
. The dissent maintains that Congress "clearly intended to avoid” the prospect that "ten-year old vehicles would be required to conform” with the emission standards. See Diss.Op. at 1580 & n. 40. However, such "ten-year old vehicles" are vehicles that were not in conformity with the emission standards during their lives, and accordingly they should not be excluded from the recall of the class of which they are mem- ’ ers. See supra at pp. 1567-1568.
Moreover, the dissent’s own proposed alternative would entail the repair of “ten-year old *1572vehicles.” The dissent would prefer that the EPA order the manufacturer to repair only vehicles and engines that were within their useful lives at the time the EPA issues its notice of nonconformity. See Diss.Op. at 1580 n. 40; id. at 1582. Under this proposal, however, manufacturers would also be required to remedy overage cars and engines at their own expense. If the administrative process and subsequent appeals took five years or more, as it has in this very case, the manufacturers would be required to "remedy decade-old cars at their own expense.” Diss.Op. at 1580 n. 40. How, then, can the dissent maintain that such a result is contrary to clear congressional intent?
. The EPA explained with specificity the purposes of section 207 when it promulgated the interpretative rule and warned that those purposes could be undermined in the absence of the rule:
The recall program has two objectives: 1) To assure that manufacturers repair vehicles which are exceeding the emission standards if maintained and used and the non-conformity occurs within the useful life of the vehicles, and 2) to encourage manufacturers to build durable emission-related components to assure that vehicles will not manifest excessive emissions during their useful lives. Both of these objectives are frustrated by an interpretation of useful life which limits manufacturers’ liability only to vehicles which are within their useful life at the time of repair.
45 Fed.Reg. at 36,397. As this court stressed only recently, ”[w]e cannot interpret section 207 ‘in a manner which runs counter to the broad goals which Congress intended it to effectuate.’ ’’ Chrysler Corp. v. EPA, 631 F.2d at 888 (quoting FTC v. Fred Meyer, Inc., 390 U.S. 341, 349, 88 S.Ct. 904, 908, 19 L.Ed.2d 1222 (1968)).
. As the EPA emphasized when issuing the interpretative rule, "interpreting section 207(c)(1) as imposing a useful life limitation on vehicles eligible for repair under a remedial plan could severely limit the number of nonconforming vehicles repaired pursuant to recall orders, could seriously impact ambient air quality, and would frustrate the intent of Congress.” 45 Fed.Reg. 36,398.
The dissent argues that the EPA could have devised an alternative administrative scheme, using its legislative rulemaking powers, to "toll” the useful life of vehicles either at the time the EPA initiates the recall or during any period of delay that could be attributed to manufacturer failure to comply in a timely manner with its statutory or regulatory obligations. See Diss.Op. at 1582-1583. Under such a scheme, only those cars within their useful lives at the time of the initial EPA notice of recall would be repaired at the manufacturer’s expense. However, under the scheme, manufacturers could not cause attrition in the class size by engaging in dilatory tactics.
We decline to enter into the inquiry of which regulatory scheme would best balance all the interests in this field. Our task here is merely to determine whether the agency's interpretation is "sufficiently reasonable,” and we conclude that it certainly is. See FEC v. Democratic Senatorial Campaign Comm., 454 U.S. 27, 39, 102 S.Ct. 38, 44, 70 L.Ed.2d 23 (1981). Regardless of whether such interpretations as the dissent suggests would be permissible — an issue we need not reach today — they are in fact not the interpretation that has been adopted by the agency. Moreover, courts should pay special deference to an agency’s interpretation when, as here, (a) the agency is interpreting a statute it is charged with administering, (b) the interpretation has been consistently adhered to, (c) Congress has acquiesced to the administrative interpretation, and (d) the statute gives the agency substantial discretion in administering and designing an enforcement scheme. See National Wildlife Fed'n v. Gorsuch, 693 F.2d 156, 166-69 (D.C.Cir.1982).