Using 2000 U.S. Census data we illustrate the importance of accounting for household specialization in lesbian couples when examining the sexual orientation gap in female labor supply. Specifically, we find the labor supply gap is substantially larger between married women and partnered lesbian women who specialize in market production (primary earners) than between married women and partnered lesbian women who specialize in household production (secondary earners). Using a semi-parametric decomposition approach, we further show that the role of children in explaining the mean labor supply gap by sexual orientation is greatly understated if the household division of labor between household and market production is not taken into account. Finally, we illustrate that controlling for children significantly reduces differences between married women and secondary lesbian earners both in terms of the decision to remain attached to the labor market (the extensive margin), as well as in terms of annual hours of work conditional on working (the intensive margin). Further, the effect of controlling for children is not uniform across the distribution of conditional annual hours; instead it primarily reduces the percentage of secondary lesbian earners working extremely high annual hours.