The vast majority of states fighting civil wars employ non-state militias in their counterinsurgent forces, and many rebel groups create similarly organized subsidiary militias to combat the government. Even after a war has ended, however, non-state militias that fought on behalf of the victor often endure. Why do some victorious militias survive longer than others in the post-war period? I argue that incumbent victors have incentives to retain the structure of their security apparatus in order to deter the resurgence of violence, but that closed autocratic victors are especially likely to retain their militias for longer periods in order to prevent possible coup attempts and a renewed rebellion. To evaluate the argument, I use data on the post-war survival and termination of victorious militias from 1989-2014. Using a survival analysis approach, I find that victorious militias survive longer in the post-war period when an incumbent closed autocracy won the war. Conversely, I find that democratic victors terminate their links with auxiliary militias quickly after the war has ended. The results have important implications for how institutions are related to security structures and the mechanisms linking militia activity and information availability to the recurrence of civil war.