The hypothesis of inequality as the source of violent conflict is investigated empirically in the context of killings by Nepalese Maoists in their People's War against their government during 1996–2003. The dependent variable is the total number of people killed during that period by Maoist rebels in each of 3,857 villages. Inequality is measured by the Gini, the Esteban-Ray polarization index, and four other between-groups indexes. Using models with district fixed effects, and instrumenting for endogeneity of the inequality measures, we find strong evidence that greater inequality escalated killings by Maoists. Poverty did not necessarily increase violence. Education moderated the effect of inequality on killing, while predominance of farmers and of Nepali speakers exacerbated it. We find evidence that more killings occurred in populous villages, lending support to the idea that violence was directed at expanding the Maoist franchise by demonstrating that opposition to the monarchy and elites in power was possible to achieve.