Exile has been the second most common fate for dictators who lost office since World War II, yet scholars know little about this phenomenon. In this article, we ask a simple yet previously unanswered question: where do exiled dictators go? We argue that three sets of factorstransnational ties, geographic proximity, and monadic characteristics of potential host statesinfluence where dictators flee. For evidence, we use original data on exile destinations to construct a directed dyadic data set of all autocratic rulers who fled abroad upon their ouster. We find that dictators are more likely to go into exile in states that are close neighbors, have hosted other dictators in the past, are militarily powerful, and possess colonial links, formal alliances, and economic ties. By contrast, fleeing dictators tend to avoid democratic states and countries experiencing civil conflict. These findings raise broader implications for several outcomes ranging from regime transitions to conflict termination.