Rescission of the Greenhouse Gas Endangerment Finding and Motor Vehicle Greenhouse Gas Emission Standards Under the Clean Air Act

The EPA received substantial comments on the proposed bases for rescinding the Endangerment Finding but relatively few specifically addressing the separate question whether we have the authority to rescind, provided that the rescission is supported by adequate grounds. Most comments received on that issue agreed that the EPA may reconsider prior actions unless the relevant statute provides otherwise and further agreed that nothing in CAA section 202(a)(1) conditions or limits our ability to reconsider prior actions. We appreciate these comments and, as noted above, are finalizing this action based on the statutory authority conferred in CAA section 202(a)(1) and the background principle that agencies may reconsider, revise, and rescind prior actions unless provided otherwise by the relevant statute. Several commenters raised contrary arguments that did not change our view from proposal. For more detailed comment summaries and responses, see the Response to Comments document.

Comment: A few adverse commenters argued that rescinding the Endangerment Finding would not support repealing the associated GHG emission standards because the standards-setting rulemakings reaffirmed and reinforced the Endangerment Finding with additional evidence. Some of these commenters also argued that CAA section 202(a)(1) is a precautionary provision, which, they asserted, means that we cannot rescind the Endangerment Finding based on a lack of confidence in the assumptions made and conclusions stated in that action.

Response: The EPA disagrees that rescinding the Endangerment Finding would not impact subsequently issued GHG emission standards and notes that these commenters misunderstand the impact of our proposal that CAA section 202(a)(1) does not authorize regulating GHG emissions in response to global climate change concerns. The Agency has consistently maintained that, at

minimum, a finding that the relevant air pollutant emissions cause or contribute to air pollution that endangers public health or welfare is a prerequisite to prescribing emission standards. In the Endangerment Finding, we asserted that the statute's “lack of specific direction” with respect to the timing of findings and of associated regulations granted “procedural discretion” to issue the actions separately. 74 FR 66501. But we maintained that the findings created the predicate authority and obligation to issue associated emission standards and acknowledged that it was at least permissible to issue the findings and standards in a single action. 74 FR 66501-02.

Finalizing the rescission of the Endangerment Finding for lack of authority under CAA section 202(a)(1) necessarily means that we lack statutory authority to prescribe or maintain GHG emission standards for new motor vehicles and engines. Whether we cited to additional evidence “reinforcing” the Endangerment Finding in subsequent rulemakings--and whether that additional evidence would itself have been sufficient to satisfy CAA section 202(a)(1) absent the Endangerment Finding--is irrelevant, as each of these actions rested on the novel statutory interpretation adopted for the first time in the Endangerment Finding. The best reading of the statute identified and applied in this final action necessarily overrides the contrary interpretation relied upon in these prior actions and therefore eliminates the legal basis for those prior actions. See section V.A and V.B of this preamble for further discussion of CAA section 202 and the legal position taken by the EPA in actions since 2009. With respect to commenters' precautionary arguments, the EPA is not finalizing the proposed alternative basis for rescission and repeal based on a new climate science finding by the Administrator. See section VI of this preamble for further discussion of the bases we are not finalizing at this time.

Comment: Some commenters argued that the CAA limits our authority to rescind prior actions, quoting NRDC v. Regan, 67 F.4th 397, 401 (D.C. Cir. 2023), for the proposition that the EPA “has no inherent authority” to reconsider its decisions. These commenters asserted that CAA section 202(a)(1) is best read as limiting our rescission authority to reconsideration under CAA section 307 or extraordinary circumstances, such as mistake or fraud, and that Congress authorized us only to update emission standards based on developments in science, technology, and economics by providing that we must “from time to time revise” emission standards “in accordance with the provisions of this section.” According to these commenters, rescinding the Endangerment Finding and associated regulations exceeds that authority.

Response: The EPA disagrees with these comments, which misconstrue the statute and misapply relevant case law. The D.C. Circuit's divided opinion in NRDC addressed our withdrawal of a regulatory determination for a drinking water contaminant under the Safe Drinking Water Act (SDWA) in lieu of issuing a national primary drinking water regulation. The panel majority and separate opinion agreed that “the power to decide is normally accompanied by the power to reconsider” unless Congress has “ `limit[ed] [the] agency's discretion to reverse itself.' ” 67 F.4th at 401 (quoting New Jersey v. EPA, 517 F.3d 574, 582-83 (D.C. Cir. 2008)). Interpreting the statutory language at issue, the panel majority concluded that SDWA section 1412 imposed such a limitation by mandating a sequential, two-step process under which the EPA “shall” propose a regulation within 24 months “[f]or each contaminant that the Administrator determines to regulate” in a final regulatory determination. Id. (quoting 42 U.S.C. 300g-1(b)(1)(A), (b)(1)(E)); but see id. at 408 (Pan, J., concurring in the judgment) (arguing that “nothing in the [SDWA] forbids the EPA from withdrawing a determination to regulate” because the “statute is silent on that issue”). NRDC did not challenge the established background principle that agencies may reconsider prior actions taken under a statutory authority absent statutory indicia to the contrary, and the language of CAA section 202(a)(1) is different in virtually every respect from the content, sequence, and timing requirements in SDWA section 1412.

CAA section 202(a)(1) sets out authority to regulate under certain conditions and provides that such regulations should be revised over time. The statutory language “from time to time revise” refers to the emission standards promulgated when the Administrator exercises “judgment” to determine that an air pollutant emitted from new motor vehicles or engines causes or contributes to air pollution which may reasonably be anticipated to endanger public health or welfare. Beyond reference to the Administrator's “judgment,” the statute contains no language constraining or limiting the power to reconsider a finding. Nor does CAA section 202(a)(1) require the EPA to establish regulations by a certain date or for certain pollutants, unlike many other provisions in CAA section 202 and throughout the CAA.\83\ Had Congress intended to restrict the repeal of CAA section 202(a)(1) emission standards based on the Administrator's findings of endangerment and contribution, it knew how to do so,as evidenced by provisions elsewhere in the statute imposing such restrictions.\84\ Additional statutory language providing that emission standards must be revised “in accordance with the provisions of this section” merely clarifies that revised standards are subject to the same conditions as the original standards (i.e., an applicable endangerment finding and the various substantive requirements for standards set out in CAA section 202(a)(2), (a)(3), et seq.). Finally, we note that this understanding of our reconsideration authority is rooted in consistent practice; as noted above, we assumed that we had such authority when denying reconsideration petitions on the merits in 2010 and 2022.

\83\ Compare 42 U.S.C. 7409 (mandating NAAQS for criteria pollutants by a date certain), 7412 (mandating regulation of hazardous air pollutants from listed source categories by a date certain), 7429 (same for waste combustors), 7521(a)(3)(B)(ii) (mandating minimum emission standards for HD vehicles for certain pollutants by a date certain), 7521(a)(6) (mandating certain control devices for LD vehicles after a date certain), 7521(b), (g)-(l) (mandating various emission standards for enumerated pollutants by dates certain).

\84\ Notably, Congress provided in CAA section 202(b)(1)(C) that the EPA cannot relax the pollutant-specific emission standards required “under [CAA section 202(b)]” when revising such standards “under [section 202(a)(1)].” 42 U.S.C. 7521(b)(1)(C). That limitation on revision authority does not apply to emission standards promulgated solely under CAA section 202(a) as an exercise of the Administrator's judgment. Comparable provisions appear elsewhere in the statute as well. See, e.g., 42 U.S.C. 7502(e) (providing that if the EPA “relaxes” a NAAQS, it must within 12 months require “controls which are not less stringent than the controls applicable to areas designated nonattainment before such relaxation”).

With respect to CAA section 307 and commenters' asserted mistake or fraud limitation, the EPA assumes commenters meant to suggest that we may only reconsider prior actions through mandatory reconsideration under CAA section 307(d) or by meeting common law standards originally developed for voiding a contract. We are not aware of any precedent establishing a mistake or fraud limitation and cannot agree that there is a plausible basis for doing so given the well-established principle that agencies may reconsider prior actions unless Congress provides otherwise. As to CAA section 307, this rulemaking followed the applicable procedural requirements set out in that provision. The mandatory reconsideration procedure in CAA

section 307(d)(7)(B) applies when a petitioner was unable to raise a centrally relevant objection during a public comment period, not to an EPA-initiated reconsideration.

Comment: A few commenters raised retroactivity concerns with the rescission and repeals, arguing that Congress must expressly authorize rules with retroactive effect and that repealing GHG emission standards for MY 2026 and earlier vehicles would be impermissibly retroactive. Some of these commenters cited Bowen v. Georgetown University Hospital, 488 U.S. 204 (1988), as setting out a clear statement rule for authority to issue retroactive rules.

Response: The EPA disagrees that repealing GHG emission standards for MY 2026 and earlier vehicles would have retroactive effect, as nothing in this final action “attaches new legal consequences to events completed before its enactment.” Landgraf v. USI Film Prods., 511 U.S. 244, 270 (1994). As a practical matter, manufacturers have already completed virtually all of the activities necessary to comply with the GHG emission standards for prior MY vehicles. Motor vehicles and engines have been designed and sold with compliant control mechanisms, the proverbial eggs are, in that sense, already scrambled. Repealing the GHG emission standards for prior MYs relieves only a limited set of compliance obligations, including certain ongoing reporting requirements, and does not impose any new or additional obligations on regulated parties.\85\ We conclude that repeal of the GHG emission standards for prior MYs is necessary notwithstanding the limited practical effect to ensure that our regulations are squarely grounded in statutory authority and avoid the inconsistency that would be created by retaining these regulations while repealing standards for future MY vehicles and engines. For further explanation of the impacts of the rescission and repeals, see section VII of this preamble and the Response to Comments document. For discussion of the distinct subject of reliance interests, see section IV.A.2 of this preamble.

\85\ For example, any contractual provisions between the seller (e.g., dealership) and a vehicle purchaser would not be changed or disrupted solely by operation of this final action.

To better assess potential reliance interests, the EPA sought comment on whether regulated parties or other stakeholders have relied in a significant and legally cognizable manner on our assertion of authority to regulate GHG emissions from new motor vehicles and engines and the requirements imposed pursuant to that asserted authority. We noted that such reliance may be relevant considerations to be weighed against competing rationales when deciding whether to change the Agency's position under relevant case law, including DHS v. Regents of University of California, 591 U.S. 1 (2020). Specifically, we sought comment on potential reliance interests by regulated parties that have expended resources complying with existing standards, including by pricing compliance into costs for consumers, and on potential reliance interests by other stakeholders on the Endangerment Finding and GHG emission standards.

With respect to regulated parties, we noted that because many compliance costs are incurred as part of research and development and during manufacturing, with the exception of the need to purchase compliance credits, this final action would have small to no impacts on MYs 2012-2024, limited impacts for MYs 2024-2026, and entirely relieve future regulatory obligations for MY 2027 and beyond. We also noted that the rescission and repeals would not mandate any particular response by regulated parties and would instead provide additional flexibility by relieving obligations. For discussion of regulatory tools available to address transitional compliance concerns, see sections III.A, VI.B, and VI.C of the preamble to the proposed rule. We also noted that regulated parties may have an interest in national uniformity and preemption and discussed the continued applicability of CAA section 209(a) and other sources of Federal preemption in sections III.A and VI.A of the preamble to the proposed rule.

With respect to other potential interests held by regulated parties and additional stakeholders, we noted that the rescission and repeals would have no impact on existing regulatory provisions for criteria pollutant and air toxics emission standards or for the separate economy and fuel-efficiency standards administered by NHTSA. We explained that general interests in regulating GHG emissions based on global climate change concerns would not justify retaining the GHG regulatory program for new motor vehicles and engines in the absence of statutory authority, and that potential dangers from exposure to the six gases combined in the Endangerment Finding would continue to be regulated when appropriate under other, more specific grants of statutory authority. For further discussion, see sections III.A and IV.A.2 of the preamble to the proposed rule. Finally, we recognized that the EPA has since relied on the Endangerment Finding as authority for GHG regulatory actions under other provisions of the CAA, including several vacated by the Supreme Court,\86\ and noted that we would address those actions as appropriate in separate rulemaking proceedings.

\86\ See West Virginia, 597 U.S. 697; UARG, 573 U.S. 302.

The EPA received significant comments on reliance interests from a variety of regulated parties and interested stakeholders that reflected diverging views on whether we should consider reliance interests, what reliance interests we should consider, and how such interests should be addressed in this rulemaking. We agree with commenters' suggestion that under Loper Bright, it is unclear how reliance interests could justify retaining or prolonging a regulatory action that is inconsistent with the best reading of the statute. Nevertheless, we carefully reviewed public comments to assess whether any aspects of this final action should be adjusted to account for reliance interests where possible to do so consistent with our statutory authority. Ultimately, we are finalizing the primary legal basis for the rescission and repeals as proposed along with the additional futility conclusions discussed above. Reliance interests raised by adverse commenters did not change our proposed view that a lack of statutory authority necessitates rescinding the Endangerment Finding and repealing the GHG emission standards and deprives us of discretion to issue revised regulations establishing a phase-out or wind-down approach. For more detailed comment summaries and responses, see the Response to Comments document.

Comment: Commenters argued that reliance interests are irrelevant when an agency proposes to rescind a prior action that exceeded its statutory authority. These commenters argued that because the EPA lacked statutory authority to issue the Endangerment Finding and associated GHG regulations, no amount of reliance could justify continuing a program that wields a power neither Congress nor the Constitution granted to the Agency. At least one commenter also cited Justice Thomas's dissenting opinion in Regents, which argued that reliance interests are irrelevant when an agency rescinds an unlawful prior action. 591 U.S. at 60.

Response: The EPA appreciates these comments and agrees that reliance

interests alone could not justify retaining or extending a regulation that exceeds our statutory authority. Particularly after Loper Bright, the relevance of reliance interests under such circumstances is unclear.\87\ On one hand, courts have consistently held that agencies must consider significant reliance interests when exercising their authority to change positions. On the other, these cases typically addressed reliance interests in contexts where the agency faced a choice between competing policy options. Under Chevron, that included the choice between permissible interpretations of the relevant statute. Now that Chevron has been overruled, however, the range of agency discretion is considerably narrowed because the best reading of the statute controls. Loper Bright, 603 U.S. at 401-04. When the statute is best read as conferring discretion, courts use ordinary tools of interpretation to “fix the boundaries of the delegated authority” and ensure the agency reasonably exercises its discretion within those boundaries. Id. at 395.\88\

\87\ Since Loper Bright, the Supreme Court has returned to the reliance interest prong of the change-in-position doctrine only in a case involving arbitrary and capricious claims that did not turn on questions of statutory interpretation. See Wages & White Lion, 604 U.S. at 567.

\88\ In Loper Bright, the Supreme Court also stated that Chevron's overruling is not a sufficient reason to invalidate “specific agency actions” upheld under the Chevron framework. 603 U.S. at 412. That stare decisis limitation does not apply to the rescission and repeals in this final action, which is a separate and subsequent decision in which the EPA is changing its interpretation of CAA section 202(a)(1) and repudiating our prior actions as exceeding our statutory authority. See, e.g., Ohio Telecom Ass'n v. FCC, 124 F.4th 993, 1002 (6th Cir. 2025) (courts are not bound by prior holdings applying the Chevron framework in the same statutory context when the agency action on review “is not the `specific agency action' ” upheld in the prior decision).

Relevant precedents decided before Loper Bright do not resolve the question whether the illegality of a prior agency action is a sufficient explanation for rescission under the change-in-position doctrine. In Encino Motorcars, LLC v. Navarro, 579 U.S. 211 (2016), for example, the Supreme Court applied the Chevron framework to an agency's decision to alter a longstanding statutory interpretation that applied an exemption to a class of employees. The Court found the change arbitrary and capricious because the agency failed to consider industry's legitimate reliance on the applicability of the exemption. Id. at 221-22. The decision appeared to assume for purposes of deciding the case that either interpretation could be permissible under Chevron and did not address whether, had the prior interpretation been unlawful, that determination would have been a sufficient explanation for the new interpretation.

In Regents, the Court found the rescission of a deferred action memorandum arbitrary and capricious for failing to consider legitimate reliance interests, even where the memorandum had provided that the deferred action program “conferred no substantive rights.” 591 U.S. at 30. That holding was informed by the Court's decision not to address whether the agency lacked statutory authority to issue the original memorandum. Compare id. at 25-28, 32, with id. at 40, 60 (Thomas, J., dissenting) (arguing that reliance interests were irrelevant because the agency was rescinding an unlawful action). Rather, the Court noted that the agency had taken the view that it retained discretion in deciding how to wind down the program, id. at 25, and assumed on that basis that the agency could have accommodated reliance interests given its “considerable flexibility in carrying out its statutory responsibility,” id. at 32.

The conclusion that we lack statutory authority under CAA section 202(a)(1) to regulate GHG emissions in response to global climate change concerns leaves us without discretion to issue revised regulations. There is no “water under the bridge” exception for unlawful agency action, and the change-in-position doctrine does not expand an agency's statutory authority for the purpose of addressing reliance interests. The Supreme Court previously rejected our efforts to reduce compliance burdens triggered by our GHG regulatory program in UARG, holding that the Tailoring Rule exceeded our statutory authority and demonstrated that the underlying Triggering Rule was itself unlawful. 573 U.S. at 328. Here, retaining or altering the GHG emission standards because of reliance interests would similarly require rewriting the statute to confer “power that neither Congress nor the Constitution” gave us. Regents, 591 U.S. at 60 (Thomas, J., dissenting). Adopting regulatory provisions to phase out or winddown the Endangerment Finding and GHG emission standards would be inconsistent with the conclusion that we lack statutory authority for the program, potentially rendering both aspects of the action arbitrary and capricious. CAA section 202(a)(1) is binary in this respect. Our authority to delay or adjust standards under additional provisions of CAA section 202 cannot be accessed without first passing through the narrow gate of CAA section 202(a)(1).

Nevertheless, as discussed below and further detailed in the Response to Comments document, we reviewed and considered reliance interests raised by stakeholders in the interest of transparency and public engagement. This discussion is not and should not be understood as a concession that such consideration is legally required, or that any disagreement with our consideration of particular reliance interests undermines this final action.

Comment: Many commenters supportive of the proposal argued that stakeholders could not have significant reliance interests warranting retention of the Endangerment Finding and GHG emission standards given the nature of the rescissions and repeals. These commenters noted that the rescission and repeals would relieve rather than impose obligations, and that manufacturers and others remain free to move forward with current plans and designs.

Response: The EPA agrees that this final action relieves compliance obligations under the CAA and does not require anything further of regulated parties with respect to GHGs. As noted at proposal, unlike the GHG emission standards, this final rescission and repeal action increases flexibility and does not require manufacturers to change plans if doing so would raise timing concerns within the MY structure of the new motor vehicle and engine market. With respect to informational labels and warranties, manufacturers may elect to proceed with implementation or not, and nothing in this final action invalidates existing labels or contracts entered into between or among manufacturers, suppliers, and purchasers. We acknowledge that regulated parties have already incurred compliance costs because of the GHG emission standards and, particularly with respect to MY 2026 and beyond vehicles, have yet to recoup such costs through sales. However, those costs were incurred because of the GHG emission standards rather than this final action and cannot legitimately be attributed to this final action. Nor is it the case that this final action deprives regulated parties of a benefit to which they would have been entitled by complying with the GHG emission standards. The “benefit” of compliance is the avoidance of enforcement actions and potential penalties under the CAA. This final action does not subject regulated parties to increased risk of enforcement.

The evaluation of reliance interests is a context-specific inquiry that turns on the structure of the regulatory program and the nature of related private

arrangements. Courts have recognized that asserted reliance interests may be unreasonable in light of the statutory scheme, Am. Fuel & Petrochemical Mfrs. v. EPA, 937 F.3d 559, 578 (D.C. Cir. 2019), and that the duty to consider reliance interests “exists in tandem with the nature of the reliance interests at issue,” Am. Petrol. Inst. v. DOI, 81 F.4th 1048, 1060 (10th Cir. 2023). CAA section 202 recognizes the MY structure of the vehicle market in various ways, including by distinguishing between “new” and existing vehicles, and we have prescribed emission standards on an MY basis for decades. Regulated parties are aware that emission standards may be changed and updated for future MYs, and, as explained above, face minimal ongoing regulatory obligations with respect to past MYs. Cases involving legally significant reliance interests by regulated parties have almost always involved agency actions that increase regulatory obligations. See, e.g., Encino Motorcars, 579 U.S. at 223. Where, as here, the agency action relieves regulatory obligations, regulated parties are not harmed by the additional flexibility of choosing between maintaining their existing plans or altering them as they see fit. See, e.g., Arizona v. EPA, 77 F.4th 1126, 1130 (D.C. Cir. 2023) (finding no standing to challenge compliance deadline extension because the rule “in no way prevented primacy states from proceeding on the original schedule”).

For these reasons, we do not believe that existing compliance investments by regulated parties are the type of significant reliance interests that warrant special consideration in the context of this rulemaking. Even taking them into account, however, such reliance interests do not expand the EPA's statutory authority under CAA section 202(a)(1). As explained above, the best reading of the statute precludes us from maintaining a GHG emission standard program for vehicles and engines. For further discussion of the bases for this final action, see section V of this preamble. For discussion of more specific compliance-related concerns, including facility investments and compliance credits, see the comment and response summaries below and the Response to Comments document.

Comment: Some commenters asserted that regulated parties have invested substantially in complying with the GHG emission standards, including by operating, constructing, and announcing facilities to manufacture EVs, and that such investments by various actors in the supply chain since 2007 amount to $211 billion. These commenters also asserted that American manufacturers have been at the forefront of developing and deploying responsive technologies, many of which are already in production and use. Several of these commenters argued that we have not justified proceeding with the rescission and repeals given these investments, while others suggested that we should consider a more limited repeal of the most recent GHG emission standards rather than a broader rescission of the Endangerment Finding.

A different set of commenters contested the relevance of such reliance interests, arguing that many of these investments predate the EPA's most recent GHG emission standards, that the most recent GHG emission standards improperly bail out automakers' bad EV investments, and that automakers are already retreating from EV production for independent reasons.

Response: The EPA acknowledges that certain regulated parties have invested significantly in EV production and technologies that have been or could be used to comply with the GHG emission standards. We also acknowledge that those companies have already reaped significant value from this program by selling credits to other companies over the years. As discussed above, however, nothing in this final action precludes market participants from continuing to make such investments or removes any benefit capable of engendering cognizable reliance interests. Nor are such investments capable of expanding the EPA's statutory authority under CAA section 202(a)(1).

In general, we do not believe that the investments in EVs and related technologies raised by commenters should be attributed exclusively to the EPA's current GHG emission standard requirements. The new motor vehicle and engine market is complex and informed by a wide variety of economic and regulatory considerations. As several commenters recognized, some of these investments predate our most recent GHG emission standards rulemakings in 2024 for MYs 2027 and beyond, and some predate the Endangerment Finding. With respect to economic influences, we note that EV demand has been subject to significant fluctuation and declines unrelated to this rulemaking. The decline in demand is attributable in part to Congress, which recently repealed certain tax credits and subsidies for EVs and disapproved three prior EPA preemption waivers for EV-forcing California vehicle and engine regulations. Changes in consumer preferences are also relevant factors. The ability of market participants to earn a return on EV and related investments thus turns on a variety of factors that ultimately fall outside the Agency's regulatory wheelhouse. The CAA requires us to take cost into account in various ways, but it does not require the EPA to ensure that EV investments turn a profit.

Comment: Several commenters asserted that automakers have relied on the EPA's GHG emission standards to export vehicles and engines overseas on the understanding that products meeting our standards will generally also meet international emission standards. These commenters argued that the rescission and repeal of U.S. GHG emission standards will create uncertainty and raise costs for regulated parties based on this additional export market concern.

Response: The EPA disagrees that possible challenges facing automakers in complying with international emission standards are legitimate reliance interests that counsel against the rescission and repeals. We question the premise that automakers assume their products will comply with applicable emission standards in export markets, as GHG emission standards are not in place for new vehicles and engines (or the same classes of new vehicles and engines) in all export markets and vary significantly among nations where such GHG emission standards are in place and applicable to imports. We also note that many automakers structure design, marketing, and production strategies to account for differing emission standards across various markets, both for GHG emissions and for emissions of criteria pollutants and air toxics. Regardless, as discussed above, nothing in this final action prevents regulated parties from maintaining current plans to the extent that they believe doing so is a convenient way to more easily participate in export markets.

Comment: Several commenters raised concerns about the GHG compliance credit regime that some regulated parties have used to comply with the existing regulations. These commenters argued that companies have accumulated credits over the past 15 years and, in some cases, already booked those credits as assets. Several of these commenters presented this as a reason not to finalize the rescission and repeals, while others requested a wind-down period.

Response: The EPA has consistently maintained that regulated parties lack a property right in compliance credits or

their use to demonstrate compliance.\89\ We note that the relevant universe of compliance credits potentially impacted by this final action is much smaller than some commenters suggest, as credits are specific to compliance years and expire after five years.\90\ Credits for MY 2020 and previous vehicles are expired, and potential credits for MY 2026 and beyond vehicles are not yet in place. These considerations lead us to conclude that the impact on stakeholders arising from compliance credit issues will be relatively small and temporary. Additionally, as discussed within the Response to Comments document, the EPA has reduced the value of emission credits within trading programs previously.

\89\ See 40 CFR 86.1865-12(k)(2) (“There are no property rights associated with CO2 credits generated under this subpart. Credits are a limited authorization to emit the designated amount of emissions. Nothing in this part or any other provision of law shall be construed to limit EPA's authority to terminate or limit this authorization through a rulemaking.”).

\90\ See 73 FR 25692 (May 7, 2010) and 40 CFR 86.1865-12(k)(2). Relatedly, see 40 CFR 86.1861-17(b)(3) (LD and MD vehicle credits); 40 CFR 1036.740(d) (HD engine credits), and 1037.740(c) (HD vehicle credits).

More fundamentally, our lack of statutory authority to retain the GHG emission standards means that we lack discretion to issue revised regulations that incorporate a phase-out or wind-down approach to address concerns related to this compliance mechanism.

Comment: Some commenters asserted that State and local governments have relied on the EPA's GHG regulatory program as a baseline to craft climate policy and invested substantial resources in EV manufacture and development, EV infrastructure, including charging stations, and transportation electrification more generally. Several of these commenters also asserted that States have relied on co-pollutant reductions from the GHG emission standards to satisfy their compliance obligations under the NAAQS for criteria pollutants. These commenters argued that, given such reliance interests, the EPA should first conclude its rescission of the Endangerment Finding, including any subsequent litigation, before repealing the associated GHG emission standards.

Response: The EPA acknowledges the comments and information received from many States and local governmental entities, including both the comments summarized above and comments from States urging us to finalize the proposed rescission and repeals. We are aware that State and local governments have, at various times, encouraged and supported the EPA's GHG regulatory program and undertaken initiatives to address perceived global climate change concerns. We disagree that this final action disrupts State and local policy initiatives that have used the Endangerment Finding or subsequent actions as a baseline, however. So long as such policy initiatives are consistent with applicable Federal law, they may continue, and nothing in this final action changes the status quo for such initiatives. To the extent commenters refer more generally to a practice of supporting and imitating aspects of the EPA's GHG regulatory program, that practice does not depend upon our continuing to maintain the program. To the extent commenters refer to information, funding, or technical support that has been integrated into such programs, we note that any such provisions are not part of the Endangerment Finding or GHG emission standards subject to rescission and repeal and that commenters did not point to a specific counterexample that should be considered in this rulemaking. Nothing in this final action addresses any separate statutory obligation the EPA may have to provide information, make grants, or provide technical support.

With respect to commenters' assertions about State and local government investments in EV technology and infrastructure, we disagree that such reliance interests counsel against the rescission and repeals for substantially the same reasons discussed above regarding regulated parties. Nothing in this final action precludes such investments, and nothing in the prior actions and rules subject to this final action entitled States or local governments to any particular benefits or return on their investments. The extent to which such investments end up supporting these entities' policy goals turns on a complex combination of unrelated regulatory and economic factors.

Finally, with respect to the NAAQS program, we note that the EPA has not established air quality criteria or NAAQS for GHGs under CAA sections 108 and 109, either individually or under the Endangerment Finding's definitional grouping of the six “well-mixed” GHGs. As explained in section VI of this preamble, this final action does not impact any of the EPA's criteria pollutant emission standards that are more directly relevant to NAAQS attainment or NHTSA's separate fuel- economy and fuel-efficiency regulations that also may result in co- benefits. We acknowledge that many regulated parties elected to comply with the GHG emission standards using technologies that also produce reductions in criteria pollutant emissions, including by shifting toward EVs or otherwise installing control equipment with co-benefits. Nevertheless, we disagree that such co-benefits engender significant reliance interests relevant to this rulemaking or that such considerations justify retaining the GHG regulatory program in the absence of statutory authority, particularly because the EPA has additional, express statutory authorities to address criteria pollutant emissions relevant to NAAQS attainment.

As a practical matter, criteria pollutant emission reductions attributable to the GHG emission standards are small in absolute terms and unlikely to materially impact States' attainment of the NAAQS. In recent GHG emission standard rulemakings, we stated our expectation that manufacturers would comply with the standards by shifting to EV production, which we predicted would lower criteria pollutant emissions from new motor vehicles, increase emissions from the power sector to accommodate additional electricity demand, and marginally decrease emissions attributed to fossil-fuel refineries given decreased demand for diesel and gasoline. For the 2024 HD GHG Emission Standards Rule, for example, we estimated small net decreases in NOX, VOCs, and sulfur dioxide (SO2) emissions and a small net increase in fine particulate matter (PM2.5) emissions.\91\ For context, the emission decreases projected for HD vehicles amount to less than 1 percent of national NOX emissions and less than 0.01 percent of VOC and SO2 emissions for 2024.\92\ As discussed above, this final action has the potential to alter vehicle emissions on a prospective basis given the MY-by-MY nature of the market and the applicability of CAA section 202(a) emission standards to “new” motor vehicles and engines. Thus, any criteria pollutant emission reductions realized in practice as a co-benefit of GHG emission standards for

MY 2025 and earlier are not impacted by this final action. Moreover, this final action does not require regulated parties to change existing plans, but rather, provides additional flexibility moving forward, meaning whether any and by how much anticipated reductions occur in practice turns on decisions by multiple independent actors.

\91\ See, e.g., 89 FR 29440, 29455 (Apr. 22, 2024).

\92\ Compare id. (estimating NOxemission reductions of 53,051 tons, VOC emission reductions of 7,272 tons, and SO2 emission reductions of 295 tons), with U.S. Environmental Protection Agency: Air Pollutant Emissions Trends Data (Apr. 2025) (estimating NOX emissions of 6,940,000 tons, VOC emissions of 12,783,000 tons, and SO2 emissions of 1,675,000 tons). National emissions are the appropriate comparator because NAAQS attainment is evaluated by criteria pollutant levels from all sources. Estimates in the 2024 HD GHG Emission Standards Rule evaluated emissions from all HD vehicles MY 2027 and beyond regardless of in-use location.

For these reasons, we cannot agree that States have significant reliance interests in the permanence of GHG emission standards in connection with NAAQS attainment. Potential impacts are limited to marginal foregone emissions reductions in future years. The co-benefits estimated in prior rulemakings are necessarily speculative because they turn on compliance decisions by manufacturers in future years and purchasing decisions by consumers (i.e., whether manufacturers comply as expected by shifting to EVs or adopting different technologies, and whether consumer demand for vehicles and engines, including relative demand for traditional vehicles versus EVs, plays out as expected). Reductions in such co-benefits are also uncertain because they depend on how regulated parties choose to proceed in future years in light of this final action. Separate and apart from this rulemaking, CAA section 202(a) makes clear that the content of the EPA's vehicle and engine emission standards are subject to revision at any time, and we have repeatedly revised the GHG emission standards for future MYs since 2010.\93\ See, e.g., Am. Fuel & Petrochemical Mfrs., 937 F.3d at 578 (finding reliance on particular biofuel volume decisions unreasonable given the EPA's express discretion to revise requirements).

\93\ Unlike CAA sections 109, 111, 112, and 129, for example, CAA section 202(a)(1) requires the EPA to revise new motor vehicle and engine emission standards “from time to time” without mandating a particular review timeline or date-certain deadline for periodic revisions. Compare 42 U.S.C. 7521(a)(1), with id. 7409(d)(1), 7411(b)(1)(B), 7412(d)(6), (f)(2), 7429(a)(5).

The appropriate mechanisms for addressing these concerns are the EPA's express statutory authorities bearing on criteria pollutant emissions and the NAAQS. We encourage States to participate in future rulemakings for criteria pollutant emission standards under CAA section 202 and other rulemakings impacting criteria pollutant emissions from stationary sources. NAAQS attainment is evaluated based on measured levels in the ambient air, and the statute provides a number of regulatory tools to the EPA and States to promote attainment. For example, the EPA may account for the impact of exceptional events and international emissions under certain circumstances and require States to adopt additional controls when their emissions contribute to nonattainment in another State. And States have discretion in formulating plans to attain the NAAQS, which may include certain mobile-source compliance programs, additional controls for new and existing stationary sources, and other emissions-reduction strategies. For additional discussion of our efforts to assist States in attaining the NAAQS, see the authorities, programs, and guidance documents referenced in the Response to Comments document.

Comment: Commenters with a variety of perspectives asserted that we failed to consider the interests of vehicle purchasers, including those with future commitments to purchase clean vehicles and past purchasers of vehicles with battery warranties and certain in-use performance requirements. Several of these commenters also stated that current GHG emission standards were projected to save consumers thousands of dollars per vehicle in fuel costs over the life of the car given continued improvements in efficiency and the availability of cleaner vehicle models, including from increased EV market penetration.

Response: The EPA disagrees that such interests counsel against finalizing the rescission and repeal and notes that commenters misconstrue the impact of this final action and the requirements in the GHG emission standards. Nothing in this final action requires regulated parties to change existing plans, and that logic applies to future purchase commitments as well. If States, municipalities, or businesses wish to fulfill existing purchase requirements or choose to purchase such vehicles in the future, they remain free to do so. Commenters provided no reason to believe that these voluntary purchase agreements were entered into to facilitate compliance with the GHG emission standards, and we are not aware of any reason that States, municipalities, or businesses not subject to the standards (i.e., not manufacturers or suppliers) would be involved in the design or production of compliance vehicles or engines. To the extent commenters meant to assert that the purchases were intended to satisfy local emission-reduction targets, many such targets are voluntary, and nothing in this final action prevents entities from proceeding with or adjusting existing strategies. With respect to past purchases, the battery warranty and in-use performance requirements cited by commenters are not set to begin until MY 2027. For this reason, purchasers cannot reasonably have relied on these requirements for past purchases, and any battery warranties or performance guarantees were entered into on a voluntary basis separately from regulatory requirements. See the Response to Comments document for additional discussion of emissions warranties and limited additional ongoing obligations for certain MY 2025 and earlier vehicles.

As to estimated fuel cost savings arising from the predicted impacts of increased market penetration of EVs, we note that fuel costs savings per vehicle for the consumer were not a substantive justification for the Endangerment Finding. Rather, we included the discussion cited by commenters in the RIAs completed for more recent standards rulemakings. Commenters did not support their contention that existing purchasers reasonably relied on the estimated fuel costs savings per vehicle from the GHG emission standards in purchasing a vehicle. Moreover, as discussed in the DRIA and RIA for this final action, we significantly adjusted prior estimates of the cost savings attributable to GHG emission standards. Our prior estimates were based on interdependent assumptions and predictions regarding future choices by unrelated actors and global fluctuations in fossil-fuel and energy supply and demand. Intervening events since our estimates in 2024, including legislative, policy, and global market changes, have already demonstrated the significant range of uncertainty inherent in the analysis. See the RIA for this final action and subsequent sections of this preamble for further discussion.

Comment: Finally, several commenters argued generally that we failed to consider reliance interests involving the U.S. economy, national security, global geopolitics, and global trade. These commenters argued that we must consider these interests to finalize a valid rule.

Response: The EPA does not believe these general assertions raise specific and legitimate reliance interests that could or must be taken into account in this rulemaking as reliance interests. Case law provides that such generalized concerns are not the type of reliance interests that require special consideration.\94\ We endeavored to take

these general concerns into account in this rulemaking when appropriate, including by carefully reviewing and considering the ways in which Congress addressed international emissions issues in the CAA. However, as discussed in section V of this preamble, the controlling statutory language in CAA section 202(a) does not authorize the Agency to regulate GHG emissions in response to such global concerns. The possibility that interpreting CAA section 202(a) to authorize regulation in response to global climate change concerns would render the statute broad enough to encompass global political and economic relations reinforces our view of the best reading of the statute.

\94\ See, e.g., Am. Petrol. Inst., 81 F.4th at 1061 (“general assertions of reliance simply do not rise to the level of ongoing and serious reliance interests necessary to trigger a duty . . . to provide a more detailed explanation”); Am. Hosp. Ass'n v. Azar, 983 F.3d 528, 540 (D.C. Cir. 2020) (rejecting general assertion of reliance interests where party “identified no reliance interests the action might be upending”).

As noted above, CAA section 202(a)(1) directs the Administrator to prescribe “standards applicable to the emission of any air pollutant from any class or classes of new motor vehicles or new motor vehicle engines, which in his judgment cause, or contribute to, air pollution which may reasonably be anticipated to endanger public health or welfare.” This core directive has remained substantially the same since Congress enacted the Motor Vehicle Pollution Control Act of 1965.\95\ Thus, a necessary condition to regulating emissions from new motor vehicles and engines is a finding--an “endangerment finding”-- that emissions of an air pollutant from a class or classes of new motor vehicles or engines cause or contribute to air pollution which may reasonably be anticipated to endanger public health or welfare.

\95\ Public Law 89-272, 79 Stat. 992, 992-93.

For the reasons discussed in sections V.A and V.B of this preamble, we are rescinding the Endangerment Finding for GHG emissions from new motor vehicles and new motor vehicle engines and, on that basis, repealing all existing GHG emission standards for passenger cars, light-duty trucks, motorcycles, buses, medium-duty vehicles, and heavy- duty vehicles and engines. The Endangerment Finding has served as the EPA's basis for regulating GHG emissions from new motor vehicles and new motor vehicle engines since 2009. Absent findings of endangerment and causation or contribution, the EPA lacks statutory authority to prescribe standards for those emissions under CAA section 202(a)(1). Thus, we must cease prescribing and enforcing standards applicable to the emission of that pollutant from new motor vehicles or new motor vehicle engines and are rescinding existing standards no longer authorized by statute.

For the reasons discussed in section V.C of this preamble, we also find that the futility of GHG emission standards for new motor vehicles and engines warrants repealing the standards separate and apart from the rescission of the Endangerment Finding. Courts have long recognized the background principle that Congress does not intend agencies to expend resources on fruitless efforts, particularly when those efforts come at the expense of express statutory obligations for which material progress is more readily achievable. Given the immense costs to manufacturers, auto workers, and American consumers, as well as the burden of administration placed on the EPA and other relevant Federal and State entities, it would be unreasonable to retain a regulatory program that does not materially further any statutory objective relevant to the global climate change concerns relied upon by the Agency in the 2009 Endangerment Finding. This conclusion is consistent with the precautionary nature ascribed by relevant court decisions to the statutory language of CAA section 202(a)(1), which we recognize does not require showing that emission standards entirely or even substantially address the identified dangers. Rather, the available information indicates that GHG emission standards have no impact at all on the adverse impacts identified in the Endangerment Finding beyond a de minimis level that falls well below inherent variability in measurements of GMST and GSLR.

Accordingly, the EPA is repealing all standards and associated test procedures adopted to limit the emission of GHGs under CAA section 202(a)(1) for highway LD, MD, and HD vehicles and engines. The EPA notes that, for LD vehicles, the Energy Policy and Conservation Act of 1975 (EPCA) \96\ and the 2007 EISA authorize NHTSA to administer the CAFE program and fuel economy labeling program. These statutes also direct the EPA to determine compliance values for manufacturers subject to the CAFE program and the fuel economy labeling program. Importantly, these statutory obligations are distinct from the EPA's authority under CAA section 202(a) and from the EPA's decisions since 2009 to regulate GHG emissions under CAA section 202(a). As explained in section VII of this preamble, we did not propose to reopen and are not finalizing in this rulemaking any changes to regulatory provisions related to our statutory roles in these NHTSA programs. Likewise, we did not propose to reopen and are not finalizing in this rulemaking any changes to criteria pollutant and air toxics standards for highway LD, MD, and HD vehicles and engines under CAA section 202(a).

\96\ Public Law 94-163, 89 Stat. 871 (1975).

In this section, the EPA provides its bases for rescinding the 2009 Endangerment Finding that initiated the Agency's unprecedented assertion of authority to regulate GHG emissions in response to global climate change concerns. Upon careful review of the text, structure, and history of CAA section 202(a)(1) and related provisions and consideration of comments received on the rationales set out in sections IV.A and V.C of the preamble to the proposed rule, we are finalizing that the Endangerment Finding and GHG regulatory program for new motor vehicles and engines exceeds the EPA's statutory authority for multiple, independent reasons. This conclusion leads us to finalize the proposed repeal of the GHG emission standards in the relevant provisions of Title 40 of the CFR as detailed in section VII of this preamble.

Section V.A of this preamble sets out our determination that CAA section 202(a) does not authorize the EPA to prescribe standards for GHG emissions based on global climate change concerns. Consistent with the Agency's practice before 2009, we conclude that this provision contains important limitations on what would otherwise be a boundless authority. First, CAA section 202(a)(1) is best read as authorizing the EPA to identify and regulate “air pollution” that threatens to endanger health and welfare through local and regional exposure. Second, CAA section 202(a)(1) is best read as requiring the EPA to apply the statutory standard for regulation as a whole by issuing findings as an integral predicate step of an emission standards rulemaking and, in doing so, evaluating whether new motor vehicle and engine emissions cause or contribute to the danger posed by the relevant air pollution. We apply the traditional tools of statutory interpretation to CAA section 202(a)(1) and related provisions, as informed by the Supreme Court's decisions in Loper Bright and UARG. We also explain how the inability of GHG emission standards to have a material (i.e., non-de minimis) impact on the dangers attributed to global climate change in the Endangerment

Finding informs our statutory interpretation.

Section V.B of this preamble explains our determination that CAA section 202(a)(1) lacks the clear congressional authorization required for the EPA to assert authority to regulate GHG emissions in response to global climate change concerns. We review the Supreme Court's precedents applying the major questions doctrine, including UARG and West Virginia, to conclude that the Nation's policy response to global climate change concerns is a question of significant economic and political importance and that Congress did not clearly empower the EPA to decide by authorizing the Administrator to “prescribe . . . standards” for emissions from new motor vehicles and engines. We further explain that a limiting construction of CAA section 202(a)(1) is necessary to avoid serious constitutional concerns with the breadth of the provision required by the logic adopted in the Endangerment Finding.

Section V.C of this preamble explains our determination, informed by comments and supporting data received in response to the proposed rule, that GHG emission standards have not and cannot materially diminish the health and welfare impacts attributed to global climate change by the Endangerment Finding in any non-de minimis way. As presented below, the results of our modeling indicate that even the elimination of all GHG emissions from vehicles in the United States (both new and existing, and inclusive of LD, MD, and HD vehicles) would not yield impacts beyond a level that is well below the range of inherent variability in measurement for trends in GMST and GSLR. We conclude that these findings lend further support to the basis for rescission in section V.A of this preamble given the language of CAA section 202(a)(1) and the background principles that Congress does not require futile efforts or include de minimis concerns in general statutory terms. We further conclude that these findings support repealing the GHG emission standards separate and apart from the rescission of the Endangerment Finding because it is unreasonable to impose immense costs that do not further any legitimate statutory purpose.

Each of the legal bases finalized in this action is separate and independent from the others, and the EPA would rescind the Endangerment Finding and repeal the GHG emission standards on any one of these bases standing alone. The EPA's lack of statutory authority for the Endangerment Finding and related regulations would require rescission and repeal even if the major questions doctrine did not apply. Similarly, the major questions doctrine would require finalizing this action even if the EPA had a plausible textual basis for asserting the authority to regulate GHG emissions in response to global climate change concerns. Each of these bases would require finalizing this action even if the futility of the GHG emission standards program were not established in the record or were not an adequate basis for this final action. Conversely, the futility of the GHG emission standards program would support repealing the GHG emission standards even if there were an adequate legal basis to retain the Endangerment Finding.

“Wisdom too often never comes, and so one ought not to reject it merely because it comes late.” Henslee v. Union Planters Nat'l Bank & Tr. Co., 335 U.S. 595, 600 (1949) (Frankfurter, J., dissenting). Because the Endangerment Finding and the regulations that rely upon it exceed the EPA's authority in multiple respects, fundamental legal principles underpinning our constitutional system compel corrective action. The Endangerment Finding must be rescinded, and the regulatory program it initiated must be, repealed.