This paper tests the broadly adopted assumption that people apply a single discount rate to the utility from different sources of consumption. Using unique data from two surveys conducted in rural Uganda including both hypothetical and real choices over different goods, the paper elicits time preferences from approximately 2,400 subjects. The data reject the null of equal discount rates across goods under a number of different modeling assumptions. These results have important theoretical and policy implications. For instance, they provide support for the idea that time-inconsistent behaviors and a corresponding demand for commitment can be observed even if individuals do not exhibit horizon-specific discounting. In addition, good-specific discounting, under certain conditions, can explain the persistence of poverty and low savings by the poor. The paper presents evidence that these conditions are satisfied in the context under study by showing that the share of expenditures on those goods with higher discount rates is decreasing with income.