OBJECTIVE The primary study aim was to examine the possible role of cadmium as a risk factor for osteoporosis by determining the bone mineral density (BMD) in workers previously exposed to cadmium. A second objective was to validate the BMD data obtained with a movable instrument. METHODS 43 workers who were exposed to cadmium for < or = 5 years before 1978 were studied. Cadmium in blood (B-Cd) and urine (U-Cd) were used as dose estimates. The BMD was assessed in the forearm, the spine, and the hip (neck and trochanter) with a dual energy x ray absorptiometry (DXA) instrument. Age and sex matched reference populations were used to compute Z scores, commonly used to assess osteoporosis. RESULTS The mean forearm Z score was -0.60 (95% confidence interval (95% CI) -1.08 to -0.12) in the group exposed to cadmium. The mean Z score for the spine was -0.47 (95% CI -0.92 to -0.03), for the hip neck -0.40 (95% CI -0.75 to -0.05), and for the hip trochanter -0.22 (95% CI -0.52 to -0.07). The decrease in forearm BMD was correlated with age (p = 0.002) and B-Cd (p = 0.040). No such correlations were found for the other sites. Workers with tubular proteinuria had a lower forearm BMD (p = 0.029) and a lower Z score (p = 0.072) than workers without tubular proteinuria. CONCLUSIONS There was a suggested dose-effect relation between cadmium dose and bone mineral density. Furthermore, there was a dose-response relation between cadmium dose and osteoporosis. Cadmium may be a risk factor for the development of osteoporosis at lower doses than previously anticipated.