Sodium fluoride stimulates bone formation and has been used to treat osteoporosis for decades despite debate about the antifracture efficacy. Hormone replacement therapy (HRT) results in only modest increases in bone mineral density (BMD). However, for women with low bone mass, the ideal therapy should not only inhibit bone resorption but simultaneously stimulate bone formation to increase bone mass above the fracture threshold. We thus performed a randomized, double-blind, placebo-controlled intervention study to prospectively investigate the effect of a low dose of fluoride, in combination with HRT, on BMD and biochemical markers of bone turnover. One hundred healthy postmenopausal women (60-70 yr old) were thus randomly assigned to: 1) HRT [transdermal 17beta-estradiol, releasing 50 microg/day; plus oral norethisterone acetate (NETA), 1 mg/day]; or 2) oral monofluorophosphate (MFP; equivalent to fluoride, 20 mg/day); or 3) HRT+MFP; or 4) placebo, for 96 weeks. All participants received a calcium supplement of 1000 mg/day. Sixty-eight women completed the study. We found a pronounced, linear increase in spinal BMD during treatment with HRT+MFP [11.8% (1.7% SEM)], which was significantly greater than the increase in the HRT group [4.0% (0.5% per yr); P < 0.05]. MFP produced a smaller increase [2.4% (0.6% per yr)], whereas there was no change in the placebo group [0.0% (0.5% SEM)]. Similar changes were found at the other skeletal sites (distal forearm, hip, and total body). Markers of bone formation showed a fall in the HRT group, which was significantly more pronounced than in the combined HRT+MFP group. A nonsignificant increase was found in the MFP group, whereas the placebo group showed a decrease caused by calcium treatment. The marker of bone resorption decreased significantly more in the HRT and the HRT+MFP groups than in the placebo group but tended to increase in the MFP group. In conclusion, this study shows, by use of biochemical markers of bone turnover, that bone resorption and formation may be dissociated, as a result of actions of two compounds with diverging effects on bone turnover. Furthermore, the synergistic effects of relatively low doses of the compounds suggested statistically and clinically significant increases in trabecular and probably also cortical bone. Adverse effects were relatively rare and mild.