Income varies considerably within countries and the locations where conflicts emerge are rarely typical or representative for states at large. Yet, most research on conflict has only examined national income averages and neglected spatial variation. The authors argue that civil conflicts are more likely to erupt in areas with low absolute income, even if a country's gross domestic product (GDP) per capita is not necessarily low, and in areas with large deviations from national averages. The authors test these hypotheses empirically using spatially disaggregated data on the location of conflict outbreaks and per capita income estimates. The authors find that areas with absolute poverty indeed see more outbreaks of conflict, and they find some evidence that inequality increases the risk of conflict. Subnational information can improve on conventional country-based measures and help our understanding of how local features and variation can give rise to mobilization and violence. (Contains 5 figures, 3 tables, and 21 notes.)