What is the political legacy of violent conflict? This paper presents evidence for a link between war, violence and increased individual political participation and leadership among former combatants and victims of violence, and uses this link to understand the deeper determinants of individual political behavior. The setting is northern Uganda, where rebel recruitment methods generated quasiexperimental variation in who became a rebel conscript and who did not. Original survey data shows that the exogenous element of conscription (by abduction) leads to significantly greater political participation later in life. The principal determinant of this increased political participation, moreover, appears to be war violence experienced. Meanwhile, abduction and violence do not appear to affect multiple non- political types of community participation. I show that these patterns are not easily explained by models of participation based on simple rational preferences, social preferences, mobilization by elites, or information availability. Only ‘expressive’ theories of participation appear consistent with the patterns observed, whereby exposure to violence augments the value a person places on the act of political expression itself. The implications for general theories of political participation are discussed.