In this paper, human capital investments are evaluated by assuming heterogeneous returns to schooling. We use the potential outcome approach to measure the causal effect of human capital investments on earnings as a continuous treatment effect. Empirical evidence is based on a sample of West German full-time employed males (BIBB/IAB survey on educational and vocational attainment and career 1998/99). Our estimate of the average partial effect (APE) of an additional year of schooling amounts to 8.7%, which is higher than OLS estimates and quite similar to conventional instrumental variable estimates.