In Sweden, children typically start compulsory school the year they turn seven. Hence, individuals born just before or just after the new year, have about the same date of birth but start school at different ages. We exploit this source of exogenous variation, to identify the effects of age at school entry on test scores, grades, educational attainment and labor market performance. Using a rich data set for the entire Swedish population born 1935-84, we find that children who start school at an older age do better in school and go on to have more education than their younger peers. Children from families with weaker educational tradition have more to win from starting school later. The long-run earnings effects are positive but small. However, since starting school later entails the opportunity cost of entering the labor market later, the net earnings effect over the entire life-cycle is negative. Exploiting within-school variation in peer age composition, we find that the school starting age effect primarily is due to absolute maturity rather than to the relative age in the class.