In South Africa, appendicular and lumbar spine bone mineral density (BMD) have been found to be similar in black and white women. However, femoral BMD has been found to be higher in black than in white women. Two different techniques were used to recalculate BMD to eliminate the possible confounding influence of ethnic differences in height on areal BMD measurements. Volumetric bone mineral apparent density (BMAD) values were calculated and bone mineral content (BMC) was corrected for body and bone size. This report analyses differences in BMD (corrected for height and weight), BMAD, BMC (corrected for body and bone size), femoral neck axis length (FNAL), mineral homeostasis and bone turnover (BT) in a group of 20 to 49-year-old premenopausal (105 whites and 74 blacks) and 45 to 64-year-old postmenopausal (50 whites and 65 blacks) female South African nurses. The corrected BMD and BMC findings were congruous, showing that both pre- and postmenopausal blacks and whites have similar distal radius and lumbar spine bone mass but that whites have lower femoral neck bone mass than blacks. In contrast, BMAD findings suggest that pre- and postmenopausal whites have lower bone mass at the lumbar spine and femoral neck than blacks but similar bone mass at the distal radius to blacks. There is a greater rate of decline in BMD in postmenopausal whites than in blacks. BMD at the femoral neck was 12.1% lower in premenopausal whites and 16.5% lower in postmenopausal whites than in blacks. There was a positive association between femoral neck BMD and weight in premenopausal blacks (R2 = 0.5, p = 0.0001) but not in whites. Blacks had shorter FNAL than whites in both the pre- and post-menopausal groups. Blacks had lower serum 25-hydroxyvitamin D (25-(OH)D) and higher 1,25-dihydroxyvitamin D (1,25-(OH)2D) levels than whites. There were no ethnic differences in biochemical markers of bone formation (serum alkaline phosphatase and osteocalcin) or bone resorption (urine hydroxyproline and pyridinoline), or in dietary calcium intake in either the pre- or postmenopausal groups. In the postmenopausal group, whites had higher ionized serum calcium (p = 0.003), similar serum albumin, lower serum parathyroid hormone (p = 0.003) and higher urinary calcium excretion (p = 0.0001) than blacks. These results suggest that the higher peak femoral neck BMD in South African blacks than in whites might be determined by greater weight-bearing in blacks and that the significantly lower femoral neck BMD in postmenopausal whites than in blacks is determined by lower peak femoral neck BMD and a faster postmenopausal decline in BMD in whites. The higher incidence of femoral neck fractures in South African whites than in blacks is probably determined by the lower femoral neck BMD and longer FNAL in whites. The greater rate of decline in BMD in postmenopausal whites than in blacks is associated with an increase in urinary calcium excretion in whites. Measurement of biochemical markers of BT has not contributed to the understanding of ethnic differences in BMD and skeletal metabolism in our subjects.