Courts, for some reason, have been addressing ultra vires challenges to agency action far more frequently in the last few years. On September 17, I ran a Westlaw search in the “All federal” database for “APA and ‘ultra vires’” that picked up 1082 opinions. Seven hundred and sixty-nine of these opinions were issued after 2020; 483 were issued after 2024. Given numbers like these, the present moment seems an especially good time to try to make sense of ultra vires review, which isn’t easy. To try to help, this post first explains how modern doctrine rests on a remarkable misreading of a leading precedent. It then briefly discusses the relatively aggressive form of ultra vires review that the D.C. Circuit has applied to the Postal Service, which seems notable in light of recent events.

Speaking of which, earlier this month, ultra vires review made a prominent appearance at the Supreme Court in Justice Alito’s dissent in United States Postal Service v. California. In that case, Alito argued that the Court should have granted the government’s application for a stay of an injunction barring implementation of a Postal Service rule governing mail-in ballots. Justice Alito contended that the plaintiffs could not invoke the APA’s cause of action to press their statutory claim as it fell within a broad exemption from APA review that Congress has granted the Postal Service at 39 U.S.C. § 410(a).  The plaintiffs therefore needed to invoke a “judicially-implied cause of action” for equitable relief, “known as an ultra vires claim.”

Quoting Justice Kavanaugh’s majority opinion in 2025’s NRC v. Texas, Justice Alito explained that this type of claim amounts to a “Hail Mary pass.” He added that a court can correct only “extreme” statutory violations where an agency has acted “entirely in excess of its delegated powers and contrary to a specific prohibition in a statute.” (Cleaned up.) Ultra vires claims must be strictly limited in this way to prevent them from providing too easy an “end-run” around statutory limits on judicial review. Previewing the merits, Justice Alito opined that plaintiffs’ “Hail Mary pass is unlikely to be successfully completed.”  

Justice Alito’s dismissive treatment of the plaintiffs’ ultra vires claim fits a common pattern. Some impediment blocks a plaintiff from invoking a statutory cause of action—e.g., the plaintiff wishes to challenge a non-final action, so the APA won’t work. To get around this problem, the plaintiff adds an ultra vires claim. The judge disposes of this claim in a paragraph explaining that ultra vires claims are almost impossible to win and gets rid of it. The Court’s opinion in NRC provides another nice example of this pattern, as well as a recent, binding precedent explaining how ultra vires review is supposed to work.

NRC traced its framework to what it called the “leading case on post-APA ultra vires review,” 1958’s Leedom v. Kyne. NRC is far from alone in this respect. Many lower court opinions have interpreted Kyne as authorizing nonstatutory review of non-final actions for egregious statutory violations. But a close read of this accidentally seminal opinion’s five paragraphs of analysis demonstrates that it did nothing of the sort—at least, not on purpose. 

Five years before enactment of the APA in 1946, the Final Report of the Attorney General’s Committee on Administrative Procedure explained that the “equity injunction has become in the United States the common remedy” in cases where plaintiffs invoke nonstatutory review to sue an officer “for conduct by the officer which, if not justified by his official authority, is a private wrong to the plaintiff.” This sounds a lot like “ultra vires review” but without the requirement of spectacular error, which was notably absent from the Final Report’s discussion as well as from contemporaneous Supreme Court opinions applying nonstatutory review, such as Stark v. Wickard. The APA, broadly speaking, codified this settled practice in 5 U.S.C. §§ 702, 704 and 706 by recognizing that any person who has suffered a “legal wrong” due to a “final agency action” has a right to judicial review to have that action “set aside.”

About a decade later, in the Kyne case, a union invoked nonstatutory review to set aside an NLRB certification order for violating § 9(b)(1) of the National Labor Relations Act by mixing professional and non-professional employees in a bargaining unit without the former’s consent. In 1940, in American Federation of Labor v. NLRB, the Court had held that such certification orders are not reviewable “final orders” within the meaning of § 10(f) of the Act, which provides that persons “aggrieved by a final order” of the Board can seek judicial review. This earlier decision did not, however, turn on whether certification orders are “final” as we use the term in administrative law—as Judge Posner would observe decades later, certification orders may not be “final orders,” but they are “orders” and they are “final” given their legal consequences. Instead, the Court’s conclusion in American Federation rested on reviewability analysis—i.e., Congress did not intend for certification orders to be reviewable under § 10(f) of the Act because allowing such review would damage the collective bargaining process.

In Kyne, the Board defended its certification order on the ground that the special statutory review scheme of § 10(f) of the Act was exclusive and precluded the district court from exercising jurisdiction. In other words, the Board was trying to extend the rule of American Federation on reviewability to preclude nonstatutory review. Rejecting this argument, the Court characterized the Act’s review scheme as limited to acts within the Board’s “jurisdiction.” It did not preclude an original suit to challenge an action that the Board had taken “in excess of its delegated powers and contrary to a specific prohibition in the Act” that was “clear and mandatory.” At bottom, the Court concluded that Congress would not want the implied preclusive force of § 10(f) to block judicial review of an action so plainly illegal as to be beyond the Board’s “jurisdiction.”

In the next few paragraphs, the Court explained that the Kyne plaintiffs could seek judicial review as the Board’s action had injured a legal right granted to them by § 9(b)(1). Definite statutory prohibitions that are “appropriate to the aim of Congress” and “capable of enforcement” should be enforced. A strong presumption of judicial review applies where its absence would mean “obliteration” of a right. Congress had “given a ‘right’ to the professional employees” that was enforceable in court.

This brief analysis confirmed that the Kyne plaintiffs could bring the bog-standard type of nonstatutory claim for equitable relief to remedy a “legal wrong” that had become common before the APA was enacted and that the APA codified. It did not suggest that nonstatutory claims, as a general matter, amount to “Hail Mary” passes or apply only to the most awful statutory violations.

Nonetheless, with the help of this unpromising material, courts of appeals, especially the D.C. Circuit, fashioned a free-floating cause of action for “ultra vires review” that can apply to spectacular statutory violations regardless of whether they are connected to a final agency action. In NRC, Justice Kavanaugh, a veteran of the D.C. Circuit (and the judge who coined the “Hail Mary” comparison), entrenched this creation in Supreme Court doctrine.

Putting provenance issues to one side, Justice Alito is certainly correct that a plaintiff invoking NRC-style ultra vires review starts out in a deep hole that will be difficult to climb out of.  It is therefore a little ironic that ultra vires review has, in the context of review of Postal Service action, sometimes been considerably more favorable to plaintiffs, providing a nice example of how doctrine evolves to fit circumstances. As Justice Kavanaugh noted in NRC, plaintiffs commonly use ultra vires review to try to evade a statutory limitation on review that might otherwise trip them up. This dynamic suggests that judges should clamp down on ultra vires review to minimize the “end-run” problem. Congress’s preclusion of APA review for most Postal Service actions, however, means that ultra vires review is often the only path available for judicial review, which creates an incentive for courts to deploy it expansively if they wish to correct legal errors and arbitrary action.

The D.C. Circuit seems to have responded to this incentive in cases like 2022’s National Association of Postal Supervisors v. United States Postal Service. In it, the Postal Service argued that an issue of statutory interpretation did not implicate the “clear and mandatory” language ostensibly required by Kyne to allow ultra vires, nonstatutory review. The court conceded that this limitation meant that it could not review claims based on “vague statutory provisions, such as one requiring an agency to use ‘appropriate data.’” Nonetheless, Kyne allowed it to reach “questions of statutory interpretation” and “questions regarding whether an agency decision was supported by a contemporaneous justification.”

It seems fair to say that, if applied, this form of Postal Service-style ultra vires review, which allows a court to reach a Chenery violation, would not pose that much of a barrier to the plaintiffs in United States Postal Service v. California. But then, the great thing about being the Supreme Court is that you get to choose your ultra vires adventure.

Richard Murphy is the AT&T Professor of Law at Texas Tech University School of Law. More details on his attempt to understand ultra vires review can be found at Ultra Vires Review of Federal Agency Action Made Simple(r), 2025 Utah Law Rev. 1201.