This paper carries out an empirical investigation on the determinants of economic growth for a group of 45 countries, including developed and developing economies. The paper focuses on the relationship between growth and income distribution. Unlike the traditional approach in which causality runs from growth to distribution, we hypothesize that greater inequality in income distribution retards a country's growth process, because greater inequality increases the possibility of social conflicti The marked contrast between Latin American and East Asian countries over the last three decades provides motivation for this hypothesis. A formal empirical analysis shows that per capita growth is influenced negatively by distributive inequality. Quantitatively, for every 10 points of increase in our measure of inequality (the ratio between the incomes of the top and bottom quintiles), per capita growth drops by 0.9% per year. Private investment is also shown to decline as income inequality increases. The results obtained here apply adequately to the entire sample of countries, which includes democracies as well as authoritarian governments. This moves beyond the theory, based on the median voter theorem, that distribution affects growth only in democracies. Furthermore, the results appear somewhat weaker when only democracies are included in the sample.