Returns to experience for U.S. workers have changed over the post-war period. This paper argues that a simple model goes a long way towards replicating these changes. The model features three well-known ingredients: (i) an aggregate production function with constant skill-biased technical change; (ii) cohort qualities that vary with average years of schooling; and crucially (iii) time-invariant age-efficiency profiles. The model quantitatively accounts for changes in longitudinal and cross-sectional returns to experience, as well as the differential evolution of the college wage premium for young and old workers.