Mortality rates are among the main indicators of the human impact of armed conflict and many surveys have assessed this impact both for targeting and evaluating humanitarian aid programmes. Almost no epidemiological analysis such as calculating relative risk was performed nor were reference values clearly described. Here the aim is to review published mortality rates for a better understanding of age-specific mortality in armed conflict. Published mortality rates from conflict situations were collected and pre-conflict reference rates composed. We calculated the relative risk of dying in conflict compared to pre-conflict for children under 5 and people older than five years old. Although limited by reporting inadequacies, the results confirm the high vulnerability of children < 5 but identify a higher relative risk of dying among the > or = 5 year olds. Although not entirely new, this observation is not fully understood. Further systematic epidemiological research is needed to estimate and understand the impact of armed conflict on mortality.