This research investigates the effect of armed conflict on adult mortality across countries and over time. Theoretical mechanisms are specified for how military violence influences adult mortality, both immediately and over time after conflict. The effects of aggregate conflict, interstate and intrastate conflicts, and conflict severity are explored. The Heckman selection model is applied to account for the conflict-induced missing data problem. A pooled analysis across 84 countries for the period from 1961 to 1998 provides broad empirical support for the proposed theoretical expectations across both genders. This study confirms the importance of both the immediate and the lingering effect of military conflict on the mortality of the working-age population. The immediate effect of civil conflict is much stronger than that of the interstate conflict, while the reverse applies to the lingering effect. Both the immediate and the lingering effects of severe conflict are much stronger than those of minor conflict. While men tend to suffer higher mortality immediately from intrastate conflict and severe conflict, women in the long run experience as much mortality owing to the lingering effects of these conflicts. The mortality data show a strong data selection bias caused by military conflict. The research findings highlight the imperative for negotiating peace. Preventing a contest from escalating into a severe conflict can produce noticeable gains in saved human lives.