Physical and lifestyle data were collected from 62 postmenopausal women who had declined hormone replacement therapy. Potential predictor variables were examined for their associations with bone mineral density (BMD) of the lumbar spine and femoral neck as assessed by dual x-ray absorptiometry. Body weight demonstrated the strongest association with lumbar BMD; lean body mass demonstrated the strongest association with femoral BMD. Together with the natural logarithm of the number of years since menopause (ln YSM) these anthropometric variables explained 36 and 34% of the variability of femoral and lumbar BMD, respectively. Serum estradiol levels demonstrated a weak positive association with BMD, which lost statistical significance after adjustment for body mass. Similarly, cardiovascular fitness was positively associated with femoral BMD prior to but not following adjustment for body mass. Controlling for years since menopause and body mass, the product of dietary calcium and calcium absorption demonstrated a weak positive correlation with femoral BMD (partial r = 0.30). The intake of tea was positively and significantly associated with both bone density measurements. In multiple regression analysis, femoral BMD was best explained by the lean body mass, ln YSM, and the daily intake of tea (r2 = 0.50). Similarly, lumbar BMD was best explained by body weight, ln YSM, and intake of tea (r2 = 0.44). Body mass is a major predictor of postmenopausal bone density at the hip and spine. A positive association between dietary calcium and BMD was detected only by taking into account the intestinal absorptive efficiency.