Epidemiologic evidence on the relation between nutrition and breast cancer is reviewed. After several decades of study, many aspects of the role of diet in breast cancer etiology are still unclear. Results from large prospective studies do not support the concept developed from animal and ecologic evidence that dietary fat intake in mid-life is associated with breast cancer risk. Thus, if fat intake is relevant to breast cancer, it is probably only at extremely low fat intakes or during early life. An emerging hypothesis that higher energy intake and growth rate in childhood and adolescence increases risk deserves further study. The possibility that diets rich in olive oil may be protective is also intriguing. Considerable evidence suggests that low intake of vegetables modestly increases the risk of breast cancer; however, the nutrients responsible remain elusive. The positive relation of alcohol intake with breast cancer risk has been seen repeatedly, and recently has been buttressed by studies showing that moderate alcohol intake increases estrogen endogenous levels. Advice to increase vegetable intake and limit alcohol consumption would probably have a modest, at best, effect on breast cancer risk. Future studies of the relation of nutrition during early life to subsequent breast cancer risk are needed.