We investigate the effects of maternity leave coverage on women?s post-birth wages, job tenure, and labor market attachment. We pay particular attention to unobservable characteristics that are correlated with maternity leave coverage and that affect labor market outcomes. We use a control sample, as well as a range of control variables, to address unobserved heterogeneity. We find evidence of substantial selection based on unobserved heterogeneity. Maternity leave effects on the three outcomes are found to be small and shortlived. This finding is understandable given that most policies in the United States allow leaves for only 12 weeks at most.