Epidemiologic evidence on the relation between nutrition and endometrial cancer (EC) is reviewed. Obesity is an important determinant of EC, probably because of its effect on the hormonal milieu of both pre- and postmenopausal women. However, epidemiologic studies of body fat distribution and EC are inconsistent, as are the data pertaining to the relation between body fat distribution and sex hormones. Randomized and observational studies of diet and sex hormones indicate that low fat diets may be associated weakly with decreased estrogen levels, and thus a lowering of EC risk. Only ecologic and case-control studies of diet and EC have been reported. These findings as well as the methodologic limitations of these study designs are discussed. Both types of studies implicate fat as a potential risk factor, while the case-control studies suggest that carotene may lower risk of EC. Epidemiologic studies of alcohol and EC also are inconsistent, but generally indicate no association, or a weak protective effect. The role of diet in the etiology of EC is unresolved. The conduct of cohort and intervention studies, which can avoid many of the methodologic shortcomings of ecologic and case-control studies, would improve our understanding of diet and EC.