OBJECTIVE To identify factors that explain a lower survival rate among black women with endometrial cancer when compared to white women. METHODS Data are from the National Cancer Institute's Black/White Cancer Survival Study, a population-based study of racial differences in cancer survival. Subjects included 329 white and 130 black women, ages 20-79 years, residing in the metropolitan areas of Atlanta, New Orleans, or San Francisco-Oakland, diagnosed with endometrial cancer from 1985 to 1987. Known prognostic factors were assessed as potential explanatory variables for the black-white survival difference using proportional hazards regression. Information was derived from interviews, abstracts of hospital and physicians' records, and a centralized review of biopsy and surgical specimens. RESULTS Adjusting for age and geographic location, risk of death among black women was 4.0 times (95% confidence interval [CI] 2.8, 5.6) that of white women. Approximately 40% of this difference could be attributed to a more advanced stage at diagnosis among black women, and 23% to tumor characteristics and treatment. Further adjustment for all remaining factors reduced the hazard ratio to 1.6 (95% CI 1.0, 2.6). CONCLUSIONS Eighty percent of the excess mortality among black women is explained by racial differences in stage at diagnosis, tumor characteristics, treatment, sociodemographic characteristics, hormonal and reproductive factors, and factors related to comorbidities and health behavior. Difference in stage at diagnosis is prominent in explaining the disparity in endometrial cancer survival rates in black and white women. Potential differences in treatment within stage merit further exploration.