At present, known risk factors account for only one-third of breast cancer cases diagnosed in the United States. They explain an even smaller fraction of the ten-fold variation in international breast cancer incidence rates. The low population-attributable risk of these identified risk factors, plus the existence of phenomena that cannot be easily explained by current etiologic hypotheses (such as the higher rate of breast cancer among black as compared to white women under age 40 within the United States), suggests that unidentified risk factors contribute substantially to breast cancer causation. This paper summarizes evidence to propose that two socially-conditioned factors determine a society's breast cancer incidence and its social gradients in risk: 1) the extent of exposure to exogenous carcinogens, and 2) breast tissue susceptibility to these exposures. It is further hypothesized that breast tissue susceptibility is inversely related to breast tissue differentiation, and that socially-mediated reproductive patterns (involving both early-terminated and full-term pregnancies) affect susceptibility both by altering (via hormonally-mediated mechanisms) the number and ratio of undifferentiated and differentiated cells, and by stimulating the growth of initiated and transformed cells. This view is presented in contrast to hypotheses that propose exposure to endogenous hormones as the major determinant of breast cancer risk.