In this paper, we investigate the role of young adult mortality on child labor and educational decisions. We argue that mortality risks are a major source of risks in returns to education in developing countries. We show that, in the absence of appropriate insurance mechanisms, the level of child labor is inefficient, but it can be too high or too low. It is too high when parents are not very altruistic and anticipate positive transfers from their children in the future. Uncertain returns to education, endogenous mortality or imperfect capital markets unambiguously increase child labor. When the level of child labor is inefficiently high, we also show that a cash transfer conditional on child's schooling can always restore efficiency regarding child labor.