The American family underwent important transformations in the last decades. Mating patterns changed, college graduates and high earners marry with each other more and more frequently. On the other hand, those at the bottom of the wage and schooling distributions have become more and more likely to stay single, and, once married or cohabiting, more likely to break up. This increasing gap in family achievements has important implications for both income and consumption inequalities, as well as intergenerational mobility. In this paper, I aim to quantify the importance of the marriage market as a channel of inequality, both at household and individual level. I build on the matching literature and set up a model of marriage, divorce and remarriage along the life- cycle in order to reproduce the afore-mentioned aggregate trends and understand the underlying drivers. In the model, risk-averse agents get married in order to benefit from joint public good expenditure, but economic gains from marriage are volatile due to labor market shocks. I show that the underlying structure of preferences and of the meeting technology are identified with matched data on the distribution of couples’ and singles’ traits, jointly with data on newlyweds and divorcees. I propose an estimation method based on indirect inference and estimate the model with PSID data. Preliminary findings suggest that differences in the productivity of household public good expenditure appear as a key driving force behind differentials in the odds of staying single, mating patterns, and, ultimately, household income inequality.