Based on administrative data for five cohorts of public school children in North Carolina, we demonstrate that those born just after the cut date for starting school are likely to outperform those born just before in reading and math in middle school, and are less likely to be involved in juvenile delinquency. On the other hand, those born after the cut date are more likely to drop out of high school before graduation and commit a felony offense by age 19. We also present suggestive evidence that the higher dropout rate is due to the fact that youths born after the cut date have longer exposure to the legal possibility of dropping out. The “crime” and “dropout” differences are strong but somewhat muted by the fact that youths born just before the cut date are substantially more likely to be held back in school. We document considerable heterogeneity in educational and criminal outcomes by sex, race and other indicators of socioeconomic disadvantage.