Evidence on the role of parental health on child schooling is surprisingly thin. We explore this issue by estimating the short-run effects of parents’ illness on child school enrollment. Our analysis is based on household panel data from Bosnia-Herzegovina, a country whose health and educational systems underwent extensive destruction during the 1992-1995 war. Using child fixed effects to correct for potential endogeneity bias, we find that – contrary to the common wisdom that shocks to the primary household earner should have more negative consequences for child education – it is especially maternal health that makes a difference as far as child schooling is concerned. Children whose mothers self-reported having poor health are about 7 percentage points less likely to be enrolled in education at ages 15-24. These results are robust to considering alternative indicators of parental health status such as the presence of limitations in the activities of daily living and depression symptoms. Moreover, we find that mothers’ health shocks have more negative consequences on younger children and sons.