To determine if differences in serum sex hormones contribute to racial differences in bone mass, independent of obesity, the authors compared bone mass and sex steroid hormone levels in 273 white women and 86 black women aged 65 years or older. This study was ancillary to the Study of Osteoporotic Fractures. Black women were recruited at two clinical centers in Baltimore, Maryland, and the Monongahela Valley, near Pittsburgh, Pennsylvania. An age-stratified random sample of white women was chosen from the Pittsburgh clinic. Demographic and historical information and anthropometric measurements were obtained from a clinic questionnaire, interview, and examination. Single photon absorptiometry was obtained at three sites: the distal radius, the proximal radius, and the calcaneus. Serum estrone, androstenedione, and testosterone were measured using extraction, column chromatography, and radioimmunoassay. Serum estrone concentrations were significantly higher and androstenedione levels were significantly lower in black women compared with white women. Racial differences in estrone could be explained largely by differences in the degree of obesity. Bone mass was greater in black women compared with white women within age, body mass index, (kg/m2), and estrone strata. Androgens were not related to bone mass. Within each race, bone mass increased linearly with increasing concentration of serum estrone. There was no interaction between race and serum estrone on bone mass. Each factor contributed independently and significantly to the regression model predicting bone mass. There was no attenuation of the effect of race if estrone was included in the model. In conclusion, race and serum estrone were independent determinants of bone mass.