We show that inequality can lead to social unrest in rural India. We develop a theoretical framework in which civilians rationally engage in social unrest to oppose the (unfair) distribution of surplus by the elite. Our model predicts that the likelihood of observing social unrest in a village increases with the sum of distances between the (log) average and the lowest incomes. We test this prediction using unique microdata on bank account information covering 2197 Indian villages. Guided by our theoretical framework, we propose an instrumental variable strategy based on the idea that positive water shocks disproportionately benefit the highest incomes in a village. Our analysis then shows that a 10% increase in our inequality measure leads to a 6.5% rise in the unconditional probability of social unrest in a given village and month.