In Sweden, children typically start school the year they turn seven. We combine this school entry cut-off with individuals' birthdates to estimate effects of school starting age (SSA) on educational attainment and long-run labour market outcomes. We find that school entry age raises educational attainment and show that postponing tracking until age 16 reduces the effect of SSA on educational attainment. On average, SSA only affects the allocation of labour supply over the life-cycle and leaves prime-age earnings unaffected. But for individuals with low-educated parents, we find that prime-age earnings increase in response to age at school start.