Civilian killings in conflict hit their highest level since 1994

A record death toll, but not from more fighting

New estimates from the Uppsala Conflict Data Program (UCDP), published in the Journal of Peace Research on 9 June 2026, put global deaths in organised violence at 244,600 in 2025.¹ UCDP's own account ranks that as the second-worst year since the 1994 Rwandan genocide, behind only 2022 — not the outright record some early coverage of the release suggested.¹⁴

The rise looks, at first glance, like a story about more war. Mostly, it isn't. Battle deaths in state-based conflict — fighting between governments and armed opposition groups, the category UCDP calls "battle-related deaths" — slipped slightly, from about 155,000 in 2024 to 153,600 in 2025.¹ Deaths in fighting between non-state armed groups, such as cartels and militias, fell too, to their lowest level since 2013.¹ Essentially all of the year's increase came from a third, distinct category: one-sided violence, UCDP's term for the deliberate killing of unarmed civilians by an organised armed actor. That toll rose 438%, from about 14,200 in 2024 to 76,500 in 2025 — on its own, more than the entire net rise in the global total.¹

One city in Darfur drove most of the global spike

Almost all of the increase traces to a single event. On 26 October 2025, the paramilitary Rapid Support Forces — by then part of a coalition called the Sudan Founding Alliance (SFA) — captured the besieged city of El Fasher in North Darfur after an 18-month siege that had trapped roughly 250,000 civilians.¹ UCDP's own assessment, built from satellite imagery and survivor interviews, put the toll at "at least 60,000" civilians killed by mid-December.¹ The UN human rights office separately documented more than 6,000 killings in the offensive's first three days alone and said the real figure was undoubtedly higher.³ British MPs were briefed that at least 60,000 had been murdered.⁵

UCDP logged one-sided violence in 32 countries in 2025 and 55 responsible armed actors, the most it has ever recorded.¹ Non-state actors, chiefly the SFA, caused 92% of those deaths; state actors caused the rest, their highest share since 2004.¹ For the first time in a decade, the single deadliest one-sided-violence actor in UCDP's data was not Islamic State but the SFA.¹ Away from Sudan, non-state violence overall kept falling — led by declines in Mexico and Brazil, where cartel fighting eased, though rising numbers of disappearances in both countries hint at possible undercounting rather than a genuine fall in lethal violence.¹

State-based fighting, the more familiar kind of war reporting, was dominated by the same conflicts as in 2024. Russia and Ukraine's war remained the world's deadliest for a third straight year, at roughly 94,700 battle deaths — about 62% of the global state-based total — though down from 102,100 in 2024 as confirmed Ukrainian losses eased even as confirmed Russian losses kept climbing.¹ Israel's war with Hamas caused an estimated 14,400 deaths, down a third on 2024 after two ceasefires. Battle deaths in Sudan's civil war itself — a separate tally from the one-sided-violence toll above — rose to about 12,200.¹

The structural picture is starker still. UCDP counted 65 active state-based armed conflicts in 2025, the most since it began tracking in 1946, including a record eight conflicts fought between states rather than within them — the most since 1987, when there were five.¹ Two of those, Russia–Ukraine and Iran–Israel, crossed the 1,000-death threshold UCDP calls a "war," the first time two interstate wars have run at once since 1987.¹ The report's own authors note the run of interstate conflict has fed a live debate among foreign-policy commentators over whether the postwar, U.S.-anchored international order is fraying — a reading the fatality data can describe but not settle.¹