The gender gap in inter-firm mobility is an important contributor to the gender pay gap but is as yet unexplained. In a structural model of workplace choice, I show that the gender mobility gap can be understood as a consequence of women's typical roles as secondary earners in most households which induces households to put more weight on the non-pay dimensions of women's workplaces. I provide direct empirical evidence for this explanation by documenting that the sensitivity of quits to wages is weaker the less an individual contributes to household earnings. Furthermore, gender differences are small once differences in earner roles are accounted for. My quantitative model evaluations show that ignoring the influence of earner roles on inter-firm mobility leads to substantial biases in wage-gap decompositions and predicted policy effects.