This dissertation addresses the argument that there may not be room for cooperation in relations with (and in between) states with authoritarian systems of rule because international conflicts may strengthen their hold on power. To this end, it asks 1) whether authoritarian regimes benefit in terms of their duration in power from conflict, 2) whether conflicts stabilize domestic politics by moderating the violence involved during regime transitions (in terms of how they fall), and also 3) whether conflicts decrease the human costs involved in the fall of an authoritarian regime. It finds that conflict does not, in general, benefit authoritarian rulers in terms of their duration in power. An adverse security environment actually makes personalist regimes more vulnerable to internal challenges and decreases their time in power. For mililtary regimes, however, conflict stabilizes domestic politics in a way that had not been systematically examined previously. A conflict-prone external environment increases the incentives for the military to act in unison and military leaders are more likely to choose a non-violent and orderly return to the barracks when facing the prospects of mass protests or armed insurgencies. This is because divisions within the military over the response to the unrest become more costly. Similarly, within an adverse security environment, the threat that wielding deadly force in an effort to retain power poses to the unity of the armed forces moderates the actions of the military and lowers the chances of force being used in ways that kill. Thus casualty levels are lower within a conflict-prone security environment for military regimes compared to those that are not.