Although sociologists have developed and tested numerous theories about suicide, economists have not analyzed this phenomenon. We derive an economic theory of suicide and test its implications using: (1) data by age in many developed countries; (2) a time series, 1947-67, by age group in the United States; (3) a cross section by state and age group in 1960. Most of our predictions are verified. Particularly striking are the response to unemployment, as strong in the postwar period as in the Depression and stronger among older individuals, and the negative effect of increased permanent incomes among all but the youngest age group, a result found in both the time series and the cross section.