On June 18, 2026, the Supreme Court issued a 5-4 decision in T.M. v. University of Maryland Medical Systems, concluding that the long-maligned Rooker-Feldman doctrine strictly bars a federal court collateral attack on a state court judgment, even when that judgment remains subject to further review in the state’s appellate courts. The Court resolved a circuit split over when the doctrine attaches and reaffirmed that litigants generally cannot use a federal district court as an alternative forum to challenge an adverse state court judgment while state appellate proceedings remain available. Although the Court described its holding as a narrow application of existing precedent, the decision reinforces the doctrine’s continued vitality.

The dispute arose after the plaintiff, T.M., was involuntarily committed to a Maryland medical center following an administrative hearing. The medical center obtained authorization to administer antipsychotic medication over her objection. T.M. and her parents responded by filing multiple state and federal actions challenging her commitment and treatment. The parties ultimately negotiated a settlement, which a Maryland state court entered as a consent order governing T.M.’s release and requiring, among other things, dismissal of the pending litigation. Ten days later, T.M., represented by new counsel, filed suit in federal district court seeking declarations that the consent order violated her due process rights and had been entered under duress, while simultaneously pursuing an appeal of the consent order in the Maryland appellate courts. Those parallel proceedings presented the question whether the federal district court had jurisdiction to entertain the challenge under the Rooker-Feldman doctrine.

Writing for the majority, Justice Sotomayor (joined by Justices Thomas, Alito, Kavanaugh, and Jackson) explained that, in the years since the Supreme Court decided Rooker and Feldman, different circuits have held that Rooker–Feldman can be triggered by state court judgments that are subject to further review in the state’s appellate courts, while others have held that Rooker–Feldman applies only where the state court proceedings that “ended.” The majority took the former view, holding that Rooker–Feldman bars suits in federal court that seek to attack a state court judgment before the highest court of that state has weighed in on it. According to the majority, the critical inquiry is not whether the judgment is sufficiently “final” for Supreme Court review under 28 U.S.C. § 1257, but whether the federal plaintiff is effectively asking a federal district court to exercise appellate review over a state court judgment—authority Congress has reserved to the Supreme Court. That, therefore, meant that the lower courts were correct to dismiss T.M.’s federal court lawsuit. The majority viewed this conclusion as “neither expand[ing] nor constrain[ing] Rooker-Feldman,” leaving it “narrowly confined” to its preexisting “strict limits.” In contrast, it characterized the “animating force behind” the dissent as “the belief that Rooker and Feldman were wrongly decided and so should be cabined whenever possible”—an issue it noted was not before the Court.

Sensing the potential demise of Rooker-Feldman, Justice Thomas wrote a concurring opinion defending Rooker as “correct as an original matter.” In his view, it correctly limited federal district courts from exercising appellate jurisdiction over state court judgments, a power limited to state appellate courts and, in some circumstances, the U.S. Supreme Court. His concurrence underscores that at least some members of the Court remain interested in defending the doctrine on originalist grounds, rather than merely adhering to precedent.

Justice Barrett dissented, joined by Chief Justice Roberts and Justices Kagan and Gorsuch. On the question presented, the dissent asserted that the proceedings in state court had not yet reached the type of final judgment contemplated by § 1257 and, therefore, were outside the scope of Rooker-Feldman and subject to review in federal court. But it also characterized Rooker-Feldman as “stand[ing] on shaky ground” and criticized the majority for expanding its application (albeit only, as the dissent put it, by “an inch”). The sharp exchange between the majority and the dissent suggests that, although the doctrine survives this round, it may very well be hanging on by a thread, with its future quite uncertain.

Takeaways

Tread carefully with Rooker-Feldman: The decision closes the door on an argument that had succeeded in some federal trial courts—that Rooker-Feldman does not bar the federal lawsuit unless and until there has been full exhaustion of state court appellate review. Going forward, parties challenging state court judgments in federal trial courts cannot avoid the doctrine simply because an appeal before the state courts remains pending. The bright-line rule is this: federal trial courts cannot review and reject state-court decisions. In other words, the ban applies to “state-court losers” whether or not they have fully exhausted their state appellate process.

At the same time, the Court repeatedly emphasized that Rooker-Feldman remains a narrow jurisdictional doctrine, not a substitute for ordinary preclusion principles or abstention doctrines. Practitioners considering dismissal under Rooker-Feldman should, therefore, continue to analyze carefully whether the federal court plaintiff is truly complaining of an injury caused by the state court judgment itself, rather than asserting an independent federal claim. In simpler terms, the Rooker-Feldman doctrine is a strict matter of subject-matter jurisdiction (whether the federal court has the power to hear the case at all), not to be confused with claim or issue preclusion (res judicata), which is an affirmative defense. Conversely, parties considering parallel federal litigation should carefully evaluate whether their claims effectively seek review or rejection of a state court judgment, even if that judgment is not yet fully final in the state appellate system.

For more information, please contact Chantel Febus, James Azadian, Andrew VanEgmond, or Sadie Betting.