There is an overwhelming tendency toward assortative matching on income and education on the marriage market. However, for women, greater income and education are often correlated with older age at first marriage, and therefore lower fertility. Using a bi-dimensional matching model with two factors, income and “reproductive capital,” I study the impact of career investments that delay marriage on women's marriage market outcomes. I find there may be an equilibrium where the highest earning women do not match with the highest earning men: these men may instead match with lowerearning, but still fertile women. The extent of this negative assortative matching on income depends on the returns to career investments for women, and the reduction in fertility linked to such investments. Women's human capital investments is in turn affected (and potentially reduced) by these anticipated marriage market outcomes. The model explains matching patterns and recent changes in marriage market outcomes for highly educated women in US Census micro data. To show that age indeed matters on the marriage market, separately from beauty, and thus that fertility acts as a form of capital, I implement an online experiment where age is randomly assigned to dating profiles. I find that men, but not women, have strong preferences for younger age, and that this preference is driven by individuals who are childless and have accurate knowledge of the age-fertility constraint.