When to Sue USCIS for Delay, and When You Will Lose

Clients ask us this constantly, and they usually ask it eighteen months into a wait, which is the wrong time. So here is the honest answer, drawn from what courts have actually done rather than from what the marketing says.

Most delay suits lose. The ones that win share identifiable features. Knowing which you have is the entire analysis.

The framework has not changed. Its application has.

Unreasonable delay claims proceed under the Administrative Procedure Act, 5 U.S.C. §§ 555(b) and 706(1), and the mandamus statute, 28 U.S.C. § 1361. Courts apply the six TRAC factors from Telecommunications Research & Action Center v. FCC.

Nothing in the past two years has modified the test. Everything has happened in how courts weigh it, and factor four is doing almost all the work.

Factor four asks the court to consider the effect of expediting delayed action on agency activities of a higher or competing priority. The government's argument is always the same: granting relief moves this plaintiff forward and everyone else back. Courts call it line-jumping, and it is winning.

What courts have held

Four years of asylum delay: not unreasonable. In Abasov v. USCIS, No. 1:25-cv-01793 (E.D.N.Y. Dec. 22, 2025), an asylum application had been pending roughly four years. Dismissed. The court found USCIS's last-in-first-out processing a "reasoned attempt to address mounting issues," noted that Congress made the 180-day asylum timeline unenforceable, and held that expediting would "simply move all others back one space."

Ten and eleven months of consular delay: not unreasonable. Liew v. Sanders, No. 1:24-cv-00342 (D.D.C. June 17, 2024): ten months in § 221(g) administrative processing, dismissed, with the observation that courts have sanctioned far longer delays. Dalmar v. Blinken, No. 1:23-cv-03315 (D.D.C. Aug. 26, 2024): roughly eleven months, dismissed, and containing the line that should govern client expectations: "immigration delays between three to five years are often not unreasonable."

EB-5 I-526 delay: not unreasonable. Da Costa v. Immigration Investor Program Office, No. 22-5313 (D.C. Cir. Aug. 18, 2023): the controlling D.C. Circuit authority. USCIS's availability-screened queue satisfies the rule of reason. Delays "undoubtedly maddening," but factor four was dispositive.

Fifty-seven months on an I-485: survived a motion to dismiss. Gao v. Mullin, No. 5:25-cv-01479-SVK (N.D. Cal. Apr. 8, 2026). This is the most useful recent decision on the plaintiff side, and its jurisdictional holding matters beyond its facts. The court rejected the government's argument that 8 U.S.C. § 1252(a)(2)(B)(ii) strips jurisdiction, drawing the distinction that keeps delay claims alive after Patel v. Garland: those cases involved final decisions; here no decision had been made. Agencies have discretion over whether to approve or deny. They do not have discretion over the timing of adjudication.

Beyond a statutory deadline: survived a motion to dismiss. Abdulraheemzai v. Noem, No. 4:25-cv-05098-JST (N.D. Cal. Apr. 24, 2026), brought by Afghan parolees challenging a USCIS memorandum that paused asylum processing entirely. Both the § 706(1) delay claim and the § 706(2) claim survived. The court held that a congressional "timetable need not be mandatory" to supply a rule of reason under factor two, the plaintiff-side counterweight to Abasov.

What the pattern tells you

Reading these together, three conclusions are defensible.

Under eighteen months is close to futile absent a statutory deadline or agency misconduct. Ten and eleven months were unremarkable to the D.D.C.

The inflection point sits somewhere past four years for discretionary benefits. Abasov rejected four years; Gao allowed fifty-seven months past the pleading stage. That gap is where the live litigation is.

A statutory deadline changes the case. Abasov failed partly because Congress made the asylum timeline unenforceable. Abdulraheemzai survived on a 150-day congressional deadline; and, per that court, the timetable need not even be mandatory.

A necessary caveat: these are individual decisions, not a dataset. We are not aware of reliable mandamus-specific filing statistics for 2025 or 2026. The immigration filing surge widely reported this year is habeas corpus driven by detention and removal, not adjudication-delay mandamus; filings in the category where delay suits sit were essentially flat year over year. Anyone telling you mandamus filings are surging is reading the wrong data.

Two things that will surprise you

The government can moot your appeal by adjudicating. In Haeri Mehneh v. Rubio, No. 25-5001 (D.C. Cir. Jan. 16, 2026), consolidated appeals by applicants stuck in administrative processing for seven and sixteen months were dismissed as moot: one applicant received a visa, the other was refused. The court never reached the merits. Filing suit is often what produces adjudication; that is frequently a win for the client and a dead end for the law.

Sue the right defendant. Liew dismissed the Secretary of State for lack of standing: the Secretary has no authority over individual visa adjudications. Standing survived only as to the Consul General and the Deputy Chief of Mission.

Where the real leverage has moved

The most significant development is not in individual delay suits at all. It is in systemic APA vacatur.

In Dorcas International Institute of Rhode Island v. USCIS, No. 1:26-cv-00132-JJM-PAS (D.R.I. June 5, 2026), Chief Judge McConnell granted summary judgment to nonprofit and union plaintiffs and vacated four USCIS policies adopted after Executive Order 14161 and the related proclamations: a global hold on asylum adjudications, a benefits hold for nationals of thirty-nine travel-ban countries, a re-review of previously approved benefits, and a policy treating country of origin as a significant negative factor in discretionary determinations. All four held contrary to law and arbitrary and capricious.

The context matters. After Trump v. CASA, Inc. (June 27, 2025) constrained universal injunctions, litigants have routed around it three ways: Rule 23(b)(2) nationwide classes, geographically limited relief, and § 706 vacatur, which holds agency action unlawful and vacates it without a formal injunction, to comparable practical effect. Dorcas is the immigration-adjudication exemplar.

The strategic implication for a client facing a delay: if the delay is the product of a policy rather than a queue, the challenge is to the policy, not to the individual pace. Those cases are winning. Individual line-jumping cases are not.

A caution on citations. Reporter citations for most 2025 to 2026 decisions above are not yet reliably available, and we cite by docket and date deliberately. Da Costa is commonly given an F.4th citation we have not independently confirmed. Nothing here should be quoted into a brief without pulling the slip opinion.

How to decide

1. Count the months honestly, against comparable cases rather than against your expectations. Under eighteen months, the answer is almost always no.
2. Look for a statutory deadline. If one exists, even a non-mandatory one, the case is materially stronger.
3. Ask whether a policy is causing the delay rather than a queue. If so, the systemic challenge is the better vehicle, and possibly one already filed.
4. Name the right defendants. Officials with actual authority over the adjudication.
5. Expect adjudication, not a ruling. Suit frequently produces a decision, sometimes a denial. Advise the client that the outcome may be an answer they do not like, delivered faster.
6. Weigh cost against realistic benefit. For a case in an ordinary queue at two years, litigation is usually the wrong instrument. Congressional inquiry, service requests, and expedite criteria cost far less.

We litigate these cases. We also tell clients when not to.