Suicide bombs in Iraq are a major public health problem. We aimed to describe documented casualties from suicide bombs in Iraq during 2003-10 in Iraqi civilians and coalition soldiers. In this descriptive study, we analysed and compared suicide bomb casualties in Iraq that were documented in two datasets covering March 20, 2003, to Dec 31, 2010--one reporting coalition-soldier deaths from suicide bombs, the other reporting deaths and injuries of Iraqi civilians from armed violence. We analysed deaths and injuries over time, by bomb subtype and victim demographics. In 2003-10, 1003 documented suicide bomb events caused 19% (42,928 of 225,789) of all Iraqi civilian casualties in our dataset, 26% (30,644 of 117,165) of injured civilians, and 11% (12,284 of 108,624) of civilian deaths. The injured-to-killed ratio for civilians was 2·5 people injured to one person killed from suicide bombs. Suicide bombers on foot caused 43% (5314 of 12,284) of documented suicide bomb deaths. Suicide bombers who used cars caused 40% (12,224 of 30,644) of civilian injuries. Of 3963 demographically identifiable suicide bomb fatalities, 2981 (75%) were men, 428 (11%) were women, and 554 (14%) were children. Children made up a higher proportion of demographically identifiable deaths from suicide bombings than from general armed violence (9%, 3669 of 40,276 deaths; p<0·0001). The injured-to-killed ratio for all suicide bombings was slightly higher for women than it was for men (p=0·02), but the ratio for children was lower than it was for both women (p<0·0001) and men (p=0·0002). 200 coalition soldiers were killed in 79 suicide bomb events during 2003-10. More Iraqi civilians per lethal event were killed than were coalition soldiers (12 vs 3; p=0·004). Suicide bombers in Iraq kill significantly more Iraqi civilians than coalition soldiers. Among civilians, children are more likely to die than adults when injured by suicide bombs. None.