Using a model that addresses the potential endogeneity of employment hours on mothers' child care time and vice versa, by including instruments based on parental attitudes, we find a significant negative (but inelastic) relation between the two time uses.