A case-control study of 168 cases with endometrial cancer and 334 controls was conducted in Birmingham, Alabama, between June 1985 and December 1988. Cases were identified at the University of Alabama Hospital and in a private practice; controls were selected from among women who attended the University optometry clinic. A food frequency questionnaire that evaluated the intake of 55 nutrients and 116 foods was obtained for 103 cases and 236 controls. Logistic regression was used to evaluate the effect of diet on endometrial cancer after adjustment for total calories, age, race, education, obesity, smoking status, age at menarche, number of pregnancies, age at menopause, diabetes, hypertension, and use of exogenous estrogens. High intake of certain micronutrients was associated with a decreased risk of endometrial cancer: the odds ratio for subjects in the upper tertile versus those in the lower tertile was 0.4 for both carotene and nitrate (95% confidence interval (CI) 0.2-0.8 and 0.2-0.9, respectively). There also was an inverse association between endometrial cancer and protein consumption (trend test; p = 0.002), and a moderate direct association with cholesterol intake (OR = 1.9, 95% CI 0.9-3.7) when terms for both these macronutrients were included in a logistic model. Total intake of animal and vegetable fat were not associated with endometrial cancer. More frequent consumption of several vegetables and certain dairy products was associated with a statistically significant decreased risk of endometrial cancer. These results suggest that diet plays an important role in the etiology of endometrial cancer.