This dissertation examines the effect formal legal institutions and preferences over those institutions have on civil war. The first chapter considers why power-sharing agreements often fail to bring lasting peace. I argue that if former rebels and incumbents have similar views regarding legal institutions being considered “legitimate,” power-sharing agreements can bring lasting peace. If, however, the former belligerents have different legal preferences, civil war is likely to recur because the legal system cannot provide a peaceful resolution of future conflicts that all parties accept. The second chapter focuses on why governments sometimes use civilian courts to try rebels rather than relying on military courts or battlefield summary justice. I argue that governments are more likely to rely on civilian courts when rebel groups heavily tax civilian populations. This is because the more a rebel group appears to be a de facto government, the more benefit the incumbent gains from delegitimizing rebels as “common” criminals. The final chapter examines why rebel groups differ in the extent to which they develop legal institutions and the effect of this variation. I argue that rebels with secessionist goals create extensive legal institutions to establish their ability to govern and that doing so lengthens the war. I conduct several large-N empirical analyses testing hypotheses drawn from these arguments, and, in each case, I find significant support for my claims.