We examine the effects of a compulsory schooling reform on child labor in Turkey, which extended the duration of schooling from 5 to 8 years while substantially improving the schooling infrastructure. We employ four rounds of Child Labor Surveys with a very rich set of outcomes. The reform reduces child labor by 4.8 percentage points (28 percent) for 12- to 17-year-olds and by 1.7 percentage points (81-percent) for 7- to 11-year-olds. For girls, the probability of spending long hours on household chores also reduces. We find that school enrollment and child labor are highly substitutable in rural areas, but not as much in urban areas. The policy effect at first increases but then sharply declines in parental income, which is consistent with the luxury axiom. Favorable effects of the reform on a large range of child labor outcomes suggest that incapacitation effects of a compulsory schooling policy (combined with investment in schooling infrastructure) can be more successful than child labor laws in combatting child labor—as monitoring school enrollment is much easier.