This paper studies how college major returns, preferences, and market design contribute to gender gaps relevant to policy. To study why gender representation gaps exist in high-earning majors, we first leverage data on major preferences to show that majors causally affect both earnings and fertility, with notable cross-field heterogeneity and modest gender differences. We then find that monetary returns largely drive men’s choice of major, while women prioritize family considerations. Our estimates suggest that 17-24% of the college gender earnings gap is due to these differing priorities, and we find suggestive evidence that variation in gender norm progressivity plays a prescriptive role in governing behavior. Finally, we evaluate policy alternatives that target representation gaps by simulating counterfactual scenarios within a centralized assignment system. Introducing small gender-based quotas modestly reduces the gender earnings gap without significantly affecting overall efficiency.