In this article in Security Studies, I challenge the diversionary theory of war. The diversionary hypothesis offers a powerful alternative to rationalist explanations of war based on the state as a unitary actor. Most recently, it has been used to explain why democratizing states are more likely to initiate the use of force. In the past two decades, however, quantitative tests have produced mixed and often contradictory empirical results regarding the relationship between domestic unrest and external conflict. This article uses a modified “most likely” case study research design to test the hypothesis. Examination of Argentina’s seizure of the Falkland Islands and Turkey’s invasion of Cyprus, two cases that should be easy for diversion to explain, provide surprisingly little empirical support for the hypothesis, raising doubts about its wider validity as well as the relationship between democratization and war.