This paper uses a unique data set with nearly career-long earnings histories to provide evidence on the returns to schooling in current and lifetime earnings. We use these results to assess the importance of life-cycle bias in earnings regressions using current earnings as a proxy for lifetime...