A confining limitation for the occupational sex segregation literature has been the inability to determine how many persons of one sex "would" have entered an occupation had the other sex not successfully entered instead. Using panel data from all American colleges of veterinary medicine (1976-1995), a fixed-effects model with lagged independent variables finds support for the "concurrent" effects of many hypothesized feminization mechanisms. Declining relative earnings and policies aimed at increasing production of graduates affect applications from men and women similarly, but feminization is driven by the decline in men's college graduation and their avoidance of fields dominated by women. The findings demonstrate the relative contributions and interdependence of supply and demand to occupational sex composition and the job search process more broadly. (Contains 3 tables, 4 figures and 7 notes.)