This paper investigates children’s time allocation to schooling, home production, and market production using a unique data set collected from northern Mali. Production shocks from harvest period pest infestations induce households to withdraw children from school and increase the probability that they are selected into farm work. Health shocks to women increases the probability that a child participates in the family business and childcare activities. These results are robust to varying assumptions about the structure of unobserved heterogeneity at the household and village levels. Different measures of household assets are also constructed to test whether assets serve as a buffer against increased child labor in response to shocks. Assets such as livestock have mixed effects on child labor and schooling, depending on the shock and asset type. However, household durables are substitutes for increased child labor when households face health shocks.