← Back to the dashboard

How we count deaths in organised violence

What the dashboard shows, where the numbers come from, how current they are, and what they can’t tell you.

The short answer

This dashboard counts people killed in organised violence: wars and other fighting in which a government takes part, fighting between armed groups such as militias, rebels or cartels, and deliberate killings of civilians. The counts come from the Uppsala Conflict Data Program (UCDP) at Uppsala University in Sweden, which has recorded every such event it can confirm since 1989.

How we know

Events. UCDP researchers read news agencies, local media and reports by the UN and NGOs, and record each event in which organised violence killed at least one person: where and when it happened, who fought whom, and a low, a high and a best estimate of the deaths. We use the best estimates.

Three types. UCDP sorts events into state-based conflict (fighting in which at least one side is a government), non-state conflict (fighting between organised groups, none of them a government) and one-sided violence (deliberate killing of unarmed civilians by a government or an organised group). On the page these are “wars and fighting involving a government”, “fighting between armed groups” and “killings of civilians”. Civilians killed in the crossfire of battles are counted under the first two.

First counts and final counts. UCDP publishes a first count, its Candidate Events data, about three weeks after each month, with quarterly files that bring each year’s first counts together. Once a year, around June, it publishes its checked final dataset for the year before. First counts grow: the first count for 2025 was 181,331 deaths, the final 245,464, mostly because many more killings of civilians in Sudan were confirmed later. So we compare this year with UCDP’s first count for the same months last year, not with the final count.

Our calculations. We add up the events by country, month and type. Events are counted in the country where they happened; UCDP records events in Gaza and the West Bank under Israel, and we show them that way. Some events can only be dated to a span of weeks or months, such as a siege or a long offensive; about a quarter of all deaths are in such events. For the monthly charts we spread their deaths evenly over the days they span, within their year, so a long battle does not show up as a spike in its first month; yearly totals are unchanged. For the deadliest conflicts we group events by UCDP’s conflict name; UCDP gives a newly coded conflict a temporary code, which we show as “not yet named”.

Our checks. On 16 November 2026 our yearly totals from UCDP’s final event data matched UCDP’s own country-year dataset exactly for every year we compared (1994, 2000, 2010, 2020, 2022, 2024 and 2025), by type of violence.

How current each part is

What the dashboard can’t tell you

Every death. Only deaths that are reported and can be confirmed are counted, so these figures are a floor. Where reporters cannot go, such as sieges and areas cut off from the internet, deaths are often found late or never.

The full cost of war. Deaths from hunger, disease and the collapse of health care caused by war are not counted, and in some wars they have been higher than the violent deaths.

Who is right. UCDP records who fought whom; it does not judge causes or blame.

Crime in general. Murders that are not part of organised violence are not counted; deaths in fighting between organised criminal groups are, as non-state conflict, once they reach UCDP’s threshold.

Further reading

Credits

Data: Uppsala Conflict Data Program, UCDP Georeferenced Event Dataset (GED) Global version 26.1 and UCDP Candidate Events Dataset version 26.0.X, licensed under CC BY 4.0. Please cite: Davies, Shawn, Therése Pettersson and Magnus Öberg (2026), Organized violence 1989–2025, and violent political protests, Journal of Peace Research; Sundberg, Ralph and Erik Melander (2013), Introducing the UCDP Georeferenced Event Dataset, Journal of Peace Research 50(4); Hegre, Håvard, Mihai Croicu, Kristine Eck and Stina Högbladh (2020), Introducing the UCDP Candidate Events Dataset, Research & Politics. UCDP is part of and funded by DEMSCORE, national research infrastructure grant 2021-00162 from the Swedish Research Council. We add up the events ourselves; the news reports UCDP cites are not reproduced. UCDP has not endorsed this site.

← Back to the dashboard