Abstract Using data from the 1995 National Survey of Family Growth, we examine residential variation in cohabiting women's union outcomes. Prior work has shown that although there are no residential differences in cohabitation, nonmetro women are more likely than others to marry directly and hold more favorable attitudes toward marriage. Building on this work, we examined residential differences in cohabiting women's union outcomes (i.e., marriage, separation, or remaining intact) to test whether nonmetro cohabiting women's unions are more likely to “end” through marriage, and whether pregnancy has a larger positive effect on marriage entry among this group. We find that nonmetro women are less likely to remain in cohabiting unions and are more likely to either marry or separate during the first 24 months of the cohabiting union. Pregnancy during cohabitation encourages marriage and discourages separation, but these effects are not significantly larger for nonmetro women.