Cross-country empirical studies have reviewed many aspects of peacekeeping missions, but the findings on their effectiveness diverge. This article draws on recent empirical literature on civil wars using a disaggregated approach, addressing the effectiveness of peacekeeping by examining the local variation in UN troop deployment and violence in the Bosnian civil war. The relationship between the intensity of local violence and troop deployment across Bosnian municipalities and peacekeeping effects on the intensity of subsequent violence are examined with a matching approach. The results indicate that although peacekeeping ‘works’, since it is deployed where the most severe violence takes place, peacekeepers have little effect on subsequent violence. This is consistent with research highlighting the obstacles to UN missions in addressing their objectives.