One of the intractable debates in the study of international conflict is the linkage between polarity and magnitude of interstate warfare. Speculations about the effects of the structure of the international system can be traced back to the Treaty of Westphalia. This article revisits this debate with a focus on war duration, which has received little attention in the literature, and presents the first theoretical discussion of the connection between polarity and war duration. It also uses a hazards model to statistically test whether five different measures of polarity are associated with war duration (1816-1992). The results provide initial support for the hypothesis that an increase in the number of poles in the state system is associated with longer wars on average. The empirical analysis and the theoretical discussion are important for understanding the consequences of the declining U.S. hegemony.