Abstract: Since commuting time is a key non-wage amenity of jobs for women (Le Barbanchon et al., 2021), public transport connectivity may be an important determinant of female employment. This paper applies a novel identification strategy to test this hypothesis. In particular, I leverage data on the rapid expansion of bus stops in the United States in 2013-2019 to study whether improved connectivity reduces the gender gap in employment. I address the canonical issue of endogeneity in public transport infrastructure development by comparing near-identical places at a super-granular level. My main identification assumption is that, within narrowly defined local labor markets, the timing and exact placement of new stops across neighboring blocks is orthogonal to changes in employment, conditional on census block fixed effects and census-tract-by-year fixed effects that absorb time invariant neighborhood heterogeneity and local labor market shocks. I find that proximity to bus stops significantly increases female employment, while male employment is largely insensitive. As a result, bus stop expansions narrow the gender employment gap, with effects that strengthen in long differences from 2014 to 2019. The evidence is consistent with a commuting channel: improved bus access increases commuting by public transport, especially for women.