How to interpret ACLED’s fatality records in Sudan

The challenge of counting deaths in Sudan

Estimating the death toll from Sudan’s civil war is exceptionally difficult. The conflict has severely limited journalists’ ability to report across the country, internet and telecommunications blackouts are frequent, and insecurity limits access for both local and international reporters. As a result, available estimates vary widely and are likely undercounts of the full death toll.

ACLED’s fatality records vs. other fatality records

Several organizations and researchers have published estimates that point to a death toll far exceeding ACLED’s figures. For example, a November 2024 study by researchers at the London School of Hygiene & Tropical Medicine estimated 61,000 all-cause deaths in Khartoum state in the first 14 months of the conflict, including over 26,000 deaths due to violence — a figure higher than ACLED's count for the entire country over the same period.¹ And as early as May 2024, US Special Envoy for Sudan Tom Perriello cited estimates of up to 150,000 people killed since the conflict began.² 

These figures reflect a broader concept of conflict mortality than what ACLED's methodology is designed to capture.

ACLED records fatalities only when deaths are reported as the immediate result of a specific incident of political violence (for more, see the ACLED Codebook). Each fatality reported must also be linked to a discrete event in order to be included in the dataset.

ACLED fatality data does not include indirect conflict-related fatalities (see FAQs: ACLED Fatality Methodology), whereas any higher-end death toll estimates for Sudan include these indirect deaths. This difference in scope is the main reason ACLED’s figures are lower than some publicly cited totals. Therefore, ACLED’s fatality figures should be interpreted as:

• A minimum documented count of deaths directly caused by political violence
• Potentially affected by underreporting due to access constraints, insecurity, and communication disruptions

How ACLED estimates fatalities in Sudan

In some exceptional cases, credible reports document a significant number of violent deaths in a given area or time period but cannot be tied to specific discrete events. In these instances, ACLED may apply “fatality splitting” — distributing the reported death toll across relevant events in the dataset based on the available contextual information. 

Taken together, ACLED's fatality figures represent a conservative estimate of fatalities, but one reached through active researcher judgment based on source reliability, available context, and the specific circumstances of each event. 

For an example of how ACLED has applied these principles in Sudan, including how updated fatality estimates from a Human Rights Watch report on West Darfur were integrated, please see this update log entry.