This paper examines the accuracy and usefulness of estimated rates of return to schooling based on the standard human capital model of Becker and Mincer. Regarding accuracy, it investigates whether failure to account for differences in ability and school quality lead to significant biases. Estimates using Ghanaian data confirm this possibility. Even when accurate, estimated rates of return to additional years of schooling may have limited policy implications; more useful for governments are rates of return to investments in school quality. The paper uses data on cognitive skills to estimate rates of return to three types of school quality improvements