Will the ceasefire between Iran and the United States continue to hold 7 days from today?

TL;DR

I estimate an 80% chance that the U.S.-Iran ceasefire will still be treated as in effect on June 5, with the main risk coming from a fresh Hormuz strike cycle before either side finalizes the 60-day extension.

Context

The forecast date is May 29, 2026, so the resolution date is June 5, 2026. I treat a ceasefire as in effect on the forecast date because the White House said on April 8 that Iran had agreed to a ceasefire and reopening of the Strait of Hormuz, a May 20 White House statement said the February 28 hostilities had terminated with the April 7 ceasefire, and AP still described a fragile ceasefire rather than resumed full-scale war on May 29 (White House, Apr. 8, White House SAP, May 20, AP, May 29).

The ceasefire is not calm. AP reported on May 29 that Trump had finished a roughly two-hour Situation Room meeting without deciding whether to approve a deal to extend the ceasefire and reopen Hormuz, while Iran said the agreement had not been finalized (AP, May 29). In the same week, the U.S. carried out self-defense strikes in southern Iran, and CENTCOM said Iran launched drones near Hormuz and a ballistic missile toward Kuwait, which CENTCOM called an egregious ceasefire violation (AP, May 25, CENTCOM, May 28).

Evidence

The historical backbone favors survival over a seven-day window. The ETH/PRIO Civil Conflict CeaseFire dataset covers 2,202 ceasefires across 66 countries and 109 civil conflicts from 1989 to 2020, and includes verbal, bilateral, unilateral, and written arrangements (PRIO). In that dataset, failed ceasefires had median durations of 65 days under a 25-battle-death threshold and 193 days under a 100-battle-death threshold (ETH/PRIO article PDF). A constant-hazard conversion gives seven-day survival of 92.8% for the 65-day median and 97.5% for the 193-day median. This is only a rough prior. The dataset is civil-war data, while this case is a direct interstate air-maritime crisis. Still, the resolution rule here is closer to a high-threshold failure measure than to an any-shot-fired measure.

A second academic anchor points the same way. Clayton and Sticher’s civil-war ceasefire-survival study, using written civil-war ceasefires from 1990 to 2019 and UCDP event data, reports that 30 days after a ceasefire starts, about 96% of definitive ceasefires, 92% of preliminary ceasefires, and 85% of cessations of hostilities remain active at a 25-fatality threshold (International Studies Quarterly, 2021). I discount that baseline sharply because the U.S.-Iran ceasefire has live blockade enforcement, direct strikes, and leader-level approval risk.

The current case has two opposing signals. The positive signal is that both sides keep using the ceasefire frame even after serious violations. AP’s explainer said the fragile ceasefire had held since April 7 despite drone and missile attacks and exchanges of fire in the Strait of Hormuz, and AP’s May 28 report said negotiators had reached a tentative 60-day extension while the latest flare-up was still being described as a wavering ceasefire, not as a resumed war (AP, May 28 explainer, AP, May 28). Axios also reported on May 28 that U.S. and Iranian negotiators had agreed on a 60-day memorandum of understanding to extend the ceasefire, pending Trump’s approval, while noting two Hormuz skirmishes in the prior 48 hours (Axios, May 28).

The negative signal is that the operating environment is still close to war. CENTCOM said on May 23 that more than 15,000 U.S. service members, more than 200 aircraft and warships, two carrier strike groups, and an amphibious ready group were supporting the blockade; it said U.S. forces had redirected 100 commercial vessels, disabled four, and allowed 26 humanitarian aid ships through over the prior six weeks (CENTCOM, May 23). IMF PortWatch chokepoint data also show that Hormuz traffic remains severely disrupted. The full comparable May history available in my extract is below; the 2026 vintage was queried on May 29 and ran only through May 24 (IMF PortWatch).

My scenario model is: 52% that the MOU is approved, announced, or effectively adopted before June 5, with a 95.5% conditional chance of YES; 30% that no MOU is signed but talks continue and the ceasefire remains ambiguously alive, with an 80% conditional chance of YES; and 18% that talks collapse or one side hardens publicly, with a 37% conditional chance of YES. That gives 80.3%: 0.52 × 0.955 + 0.30 × 0.80 + 0.18 × 0.37 = 0.8032. I put the MOU path above even odds because AP and Axios say negotiators reached tentative terms, but not much above even odds because AP says Trump did not decide after the May 29 meeting and Iran said the agreement was not final (AP, May 29, Axios, May 28).

What's non-obvious

The ceasefire can survive more shooting. That is the key point. The question’s threshold is not no violations; it is no acknowledged resumption of sustained or large-scale U.S.-Iran hostilities. The May 25 U.S. strikes and the May 27-28 Iranian missile and drone episode would look like breakdowns under a stricter definition, but AP, CENTCOM, and the diplomatic reporting still treated them as violations inside a fragile ceasefire rather than as the end of the ceasefire (AP, May 25, CENTCOM, May 28, AP, May 28 explainer).

The same fact also caps the forecast. Hormuz is both the brake and the trigger. The traffic collapse gives both sides a reason to find a face-saving extension, but the blockade puts U.S. forces, Iranian missiles, drones, mines, and commercial vessels in daily contact. That makes another isolated incident likely and a resolution-relevant incident plausible. I therefore stay near 80%, not the high 80s.

Limitations

The biggest limitation is that the public MOU text was not available at the forecast time. AP reported on May 29 that the meeting ended without a Trump decision, that Iran said no agreement was finalized, and that nuclear, Hormuz, Hezbollah/Lebanon, and frozen-funds issues remained live (AP, May 29). A signed MOU would move this forecast up fast; a public rejection by Trump or Tehran would move it down fast.

The base-rate data are imperfect. The best large ceasefire dataset I found is civil-conflict data from 1989 to 2020, not a direct interstate-war dataset for nuclear-adjacent maritime crises involving the United States and Iran (PRIO). The PortWatch data are useful as a pressure gauge, but they are lagged through May 24 in the extract I used, so they cannot tell us whether the May 29-30 military situation is improving or worsening (IMF PortWatch). The final resolution will depend less on the fact of another shot and more on whether AP, Reuters, BBC, the Washington Post, and the New York Times describe events as the ceasefire ending.