I study the effects of the distribution of individual abilities on economic growth using a Lucas (1988) type two-sector growth model. The relation between investment in education and individual education efficiency is positive but less than one-to-one, so the model predicts a negative relation between the variance of ability and aggregate investment in education via Jensen’s inequality theorem. As a result, a more homogeneous country is predicted to achieve a higher growth rate than a more heterogeneous country within finite time. This result obtains in the absence of any political mechanism commonly used in the inequality-growth literature. This novel feature remains robust to a convex education technology of Rebelo (1991).