The relationship between diet and breast cancer has been analyzed by animal, ecologic, migrant, and epidemiologic studies. The 14 cohort and 33 case-control studies that have been published to date are reviewed in this article. Factors considered in these studies include caloric intake, as well as fat, protein, fiber, beta-carotene, and vitamin E and C consumption. The results of the published studies are summarized, and the point estimates of the risks corresponding to the highest category of consumption as compared to the lowest are presented in figures. Some of the disagreements among studies could be explained by the methodologic difficulties inherent in dietary investigations, such as the establishment of an accurate dietary history, or insufficient diversity in exposure. Further studies taking these points into account and minimizing biases inherent to a case-control design might help to elucidate the relationship between diet and breast cancer, and to define dietary recommendations. Only large long-term cohort studies such as are now in progress can help to resolve the still unanswered questions concerning the contribution of these dietary factors to the risk of breast cancer. We suggest the establishment of new dietary cohorts and the continued follow-up of the existing cohorts.