This paper investigates to what extent assortative mating contributes to intergenerational earnings persistence. I use an errors-in-variables model to demonstrate how pooling of partners' "potential" earnings affects intergenerational earnings persistence, and simulate persistence under different assumptions about assortative mating and women's earnings distribution. Using Swedish data on cohorts born 1945-1965, I show that a substantial decline in marital sorting has contributed little to lowering intergenerational persistence. Variations in marital sorting must be large to affect intergenerational mobility to a great extent. Instead, the relative earnings distributions of men and women, in combination with sorting, are important for intergenerational persistence.