Thirty-one of 62 consecutive premenopausal women with primary cancer of the breast completed a 1-year investigation period, receiving either 30 mg tamoxifen daily (15 patients) or placebo (16 patients). They were examined at the operation (t0) and at 3-month intervals (t1, t2, t3, and t4). Bone mineral content (BMC) was measured at operation and after 12 months. Fifty-six patients with benign tumors were included as healthy controls. All values of both cancer treatment groups and of the benign tumor group were comparable at the time of operation. BMC decreased significantly in both cancer patient groups when 12-month values were compared to the initial level (tamoxifen, -3.2%, P less than 0.001 and placebo -2.5%, P less than 0.01). However, no significant difference in BMC changes was noted between tamoxifen and placebo treatment. The serum phosphate was significantly decreased in the tamoxifen treatment group at each examination. In the placebo group, the alkaline phosphatase level increased significantly at each examination, whereas the serum magnesium fell at the 6- and 12-month examinations. All other biochemical indices of calcium metabolism were basically unchanged. It is concluded that BMC is reduced in metastatic breast cancer through osteolytic metastatic bone foci. Tamoxifen also decreases the BMC. It is, however, unclear if this effect is due to a progression of the disease in spite of the treatment or if it is caused by a direct action of tamoxifen on bone.