We analyze the way women's education influences the effect of children on their level of labor market involvement. We propose an econometric model that accounts for the endogeneity of labor market and fertility decisions, for the heterogeneity of the effects of children and their correlation with the fertility decisions, and for the correlation of sequential labor market decisions. We estimate the model using panel data from NLSY79. Our results show that women with higher education work more before the birth of the first child, but children have larger negative effects on their level of labor market involvement. Differences across education levels are more pronounced with respect to full time employment than with respect to participation. Other things equal, higher wages reduce the effect of children on labor supply. Controlling for wages, women with higher education face larger negative effects of children on labor supply, which suggest they are characterized by a combination of higher marginal product of time spent in the production of child quality and higher marginal product of time relative to the marginal product of other inputs into the production of child quality.