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

TL;DR

I estimate a 63.8% chance the U.S.–Iran ceasefire will still be judged in effect on June 10, 2026, because major sources still describe an operative but badly violated ceasefire rather than an acknowledged return to sustained war.

Context

The forecast date is June 3, 2026, so the resolution date is June 10, 2026. The April ceasefire still exists in the resolution sense: the White House said on April 8 that Iran had agreed to a ceasefire, the May 20 administration position said February 28 hostilities had terminated with the April 7 ceasefire, and current Reuters/AP/CENTCOM language still uses “shaky,” “fragile,” or “ongoing” ceasefire framing.

The ceasefire is not quiet. On June 2–3, Iran fired missiles and drones toward Kuwait and Bahrain, Kuwait said drones hit its main airport and killed one person, and U.S. forces struck a ground-control station on Qeshm Island. AP called this a test of a fragile ceasefire; Reuters said a shaky ceasefire was still in place. That language is the core reason this is not already near zero.

Evidence

The historical prior says a ceasefire that has survived almost two months usually survives one more week, but this prior is weak. The ETH/PRIO Civil Conflict CeaseFire dataset covers 2,202 civil-conflict ceasefires from 1989–2020 across 66 countries and 109 conflicts; among failed ceasefires, the median duration was 65 days using a 25-fatality threshold and 193 days using a 100-fatality threshold. A constant-hazard conversion gives one-week survival of about 93% at the 65-day median, but this case is a direct U.S.–Iran interstate confrontation under blockade, not a civil-war truce.

The more useful base rate is this ceasefire’s own behavior. Since late May it has absorbed repeated direct kinetic incidents without major sources calling it over. On May 28, CENTCOM called an Iranian missile toward Kuwait and five drones near Hormuz an “egregious ceasefire violation,” not a ceasefire collapse. On May 31, CENTCOM said U.S. strikes on Goruk and Qeshm followed Iran’s shootdown of a U.S. MQ-1, but described the response as occurring during the “ongoing ceasefire.” On June 2, CBS reported further U.S. strikes on Qeshm after Iranian missile and drone attacks, and said U.S. forces maintained that the ceasefire remained in place.

The current incident tempo is still dangerous. Reuters reported on June 3 that Iranian attacks on Bahrain, Kuwait, and other regional targets were thwarted or failed, that U.S. forces downed drones and struck Qeshm Island, and that diplomacy showed little progress. AP reported the same cycle, plus the Kuwaiti claim that Iranian drones damaged Kuwait International Airport, killed one person, and wounded dozens. Washington Post called the exchange one of the most intense bouts since the ceasefire began in April.

Shipping data show why the crisis has not normalized. IMF PortWatch daily chokepoint data run from March 2023 through May 31, 2026; the latest vintage available here has 1,000 daily observations for the Strait of Hormuz. The units are daily AIS-derived vessel transits and estimated cargo capacity, not seasonally adjusted. The full annual history in that series is:

May 2026 was far worse than the year-to-date average: 4.7 vessels per day versus 105.5 in May 2024 and 106.5 in May 2025. That supports the news picture: Hormuz remains both the main bargaining chip and a daily accident generator. AP reported that a tentative 60-day ceasefire extension would address Hormuz by forbidding tolls and requiring mine removal within 30 days.

Diplomacy keeps the YES probability above 50%. AP reported on May 28 that U.S. and Iranian negotiators had reached a tentative agreement to extend the ceasefire by 60 days and start new nuclear talks, though Iran had not confirmed the deal and Vice President JD Vance said Trump had not yet approved it. Reuters reported on June 3 that Trump said talks were continuing continuously, while Iranian media said communication had paused. AP also reported Trump saying Israel’s fighting with Hezbollah was holding back Iran talks, which implies the White House still wants the Iran ceasefire framework to survive.

U.S. domestic politics also lean toward keeping the ceasefire label alive. AP reported that the House approved a war-powers resolution on June 3 by 215–208 to halt U.S. military action against Iran, with four Republicans joining Democrats; the resolution may not bind Trump, but it raises the political cost of an acknowledged return to major war.

My estimate uses a scenario tree. I put a 96% chance on the ceasefire being treated as still in effect at the moment of forecast, despite the June 2–3 exchange. Conditional on that: a 26% chance of a formal extension or clear diplomatic reaffirmation, with 95% YES; a 40% chance talks limp on and incidents stay bounded, with 80% YES; a 22% chance of another severe direct exchange but no consensus that the ceasefire ended, with 42% YES; and a 12% chance of an official or consensus breakdown, with 4% YES. This gives $0.96 \times [0.26(0.95)+0.40(0.80)+0.22(0.42)+0.12(0.04)] = 0.638$.

What's non-obvious

The obvious read is that a ceasefire cannot include missiles, drones, and airstrikes. But this question resolves on acknowledged resumption of sustained or large-scale direct hostilities. Major reporting has not crossed that line. Reuters still says a shaky ceasefire is in place; AP says the attacks test a fragile ceasefire; CENTCOM says it is acting during an ongoing ceasefire. The ceasefire has become a political container for limited direct violence.

The second non-obvious point is that violations can make the ceasefire both weaker and stickier. Repeated exchanges raise the chance of a casualty or misfire that forces reclassification. But they also show a high tolerance for bounded violence under the ceasefire label. The sharpest NO path is not just another strike. It is a strike that kills U.S. personnel, hits a U.S. warship or major base, causes a large Iranian casualty event, or triggers an official statement that talks and the ceasefire are over.

Limitations

The largest uncertainty is definitional. If resolvers treat the June 2–3 missile/drone/strike cycle as already amounting to “renewed active hostilities,” the probability should be far lower. I do not because the major sources most likely to govern resolution still frame the situation as a ceasefire under severe strain, not a ceasefire that has ended.

The historical data are only a loose guide. The best large-N ceasefire datasets are mostly civil-conflict data with fatality thresholds, while this is a direct U.S.–Iran interstate conflict with a naval blockade and linked Israel–Hezbollah fighting. The PortWatch series is useful but lags the forecast date: the latest data point I could verify was May 31, three days before this forecast. Private facts would move the forecast most: whether Trump has decided to sign the 60-day MOU, whether Iranian commanders are authorized to keep firing around U.S. bases, and whether Israel expands operations in Lebanon before June 10.