Elite female ballet dancers exhibit several risk factors for osteoporosis during their performing years. To study the long-term effect of this lifestyle, we compared the bone mineral density (BMD) of 101 retired elite female ballet dancers (mean age 51 years, SD = 14 years) and 101 normal controls, derived from a twin study, matched hierarchically for age, height, weight, and menopausal status. The dancers, who had been retired for a mean of 25.6 years (range 1-53 years) reported a greater prevalence of previous menstrual disturbance, greater lifetime alcohol intake and smoking, and a lower dietary milk intake in adolescence than controls (all p < 0.05). However, current exercise in the dancers was twice that of the controls (p < 0.01). The BMD of retired dancers did not differ from that of the controls at weightbearing sites. The mean +/- SE difference in BMD (dancers minus controls) was 0.009 +/- 0.013 at the total body, -0.009 +/- 0.018 at the total hip, 0.005 +/- 0.017 at the femoral neck, 0.014 +/- 0.018 at the femoral trochanter, 0.036 +/- 0.022 at the femoral intertrochanter and -0.017 +/- 0.021 at the lumbar spine. Retired dancers had lower mean (+/- SE) BMD at the nonweightbearing sites: ultradistal radius (-0.029 +/- 0.008) (p < 0.01) and at the midthird radius (-0.019 +/- 0.011) (p < 0.05). There was no difference in the proportion in each of the World Health Organization (WHO) categories of osteopenia (t score -1.0 to -2.5) and osteoporosis (t score < -2.5) at any of the measured sites. Regression analysis revealed that menstrual disturbance was negatively associated with BMD at the lumbar spine and the ultradistal radial sites, but not at the weightbearing femoral sites. This sample of retired elite ballet dancers who had multiple historical risk factors does not appear to have an increased risk for future hip or vertebral fracture based on WHO standards.