When did marriage become strongly assortative? I use a uniquely suitable database from Quebec 1800–1970 to provide the long-run perspective necessary to answer this question. First, I develop a novel method that reveals that marriage was highly assortative as far back as the early nineteenth century. Next, I show this matching depends on the individual human capital of women, not just on family backgrounds. Finally, I show that mothers had an effect on child outcomes independent of the fathers. Thus, despite deeply conservative gender norms, marriage matching—and women—have always mattered for social mobility.