Bone mineral content (BMC), bone mineral density, and metacarpal dimensions were studied in 50 women with Turner's syndrome aged 21-45 years in relation to karyotype, estrogen treatment, physical fitness, and biochemical markers of bone turnover. No differences were found between the 25 women with karyotype 45.X and women with other karyotypes. Forty-six women had received estrogen. Significant partial correlations were found between bone mineral density of the forearm and duration of estrogen treatment and physical fitness. BMC of the lumbar spine corrected for vertebral height (BMC(C)spine) was directly correlated with duration of estrogen treatment and height, marginally correlated with physical fitness, and inversely correlated with age. Outer metacarpal width was positively correlated with duration of estrogen treatment, age at initiation of therapy, and body weight. The diameter of medullary space showed negative correlation with physical fitness and height, and positive correlation with age at initiation of estrogen treatment. Cortical thickness was positively correlated with duration of estrogen treatment, physical fitness, and height. No convincing effects of estrogen could be demonstrated in women below the age of 30. Above the age of 30, all bone mineral measurements were markedly elevated in women treated for longer than the average of this age group. BMC(C)spine was inversely correlated with biochemical markers of bone formation. Our results demonstrate that estrogen treatment and physical fitness are important determinants of bone mineral status in Turner's syndrome and add to the evidence that estrogen treatment increases BMC in Turner's syndrome.