The proportion of U.S. high school students working during the school year ranges from 23% in the freshman year to 75% in the senior year. This study estimates how cumulative work histories during the high school years affect probability of dropout, high school academic performance, and the probability of attending college. Variations in individual date of birth and in state truancy laws along with the strength of local demand for low-skill labor are used as instruments for endogenous work hours during the high school career. Working more hours during the academic year does not affect high school academic performance. However, increased high school work intensity raises the likelihood of completing high school but lowers the probability of going to college. These results are similar for boys and girls, and so working during high school does not explain the widening gap in college entry between men and women.