This paper analyses the effect of child labor on household labor supply using 1920 US Census micro data. The aim of the analysis is to understand who in the household benefits from child labor. In order to identify a source of exogenous variation in child labor I use State-specific child labor laws. I find that a rise in the proportion of working children by household is associated with no variation in parents' labor supply. I also find a strong negative externality among children: as the proportion of working children by household rises, everything else equal, the probability that each child works falls while the probability that he attends school rises. This suggests that parents redistribute entirely the returns from child labor to the children in the household, consistent with a model of household labor supply with fully altruistic parents.