Violent crime in Colombia is analyzed following the demobilization of ex-combatants using municipal-level data. The main findings are that an increased presence of ex-combatants does not systematically increase homicides, but may increase robberies. Reintegration programs are shown to matter. Former paramilitaries who are not in a reintegration program increase crime. Former guerillas increase robberies, regardless of whether they are in or out of reintegration, but homicides decrease for guerrillas in reintegration. Ex-combatants often settle in municipalities with more crime. Controlling for reverse causality, ex-combatants only increase crime if they are not in reintegration, while in reintegration they may reduce crime.