In order to assess the relative roles of the androgenic and/or estrogenic components in the stimulatory effect of dehydroepiandrosterone (DHEA) on bone mineral content (BMC) and density (BMD), ovariectomized (OVX) female rats received DHEA administered alone or in combination with the antiandrogen flutamide (FLU) or the antiestrogen EM-800 for 12 months. We also evaluated, for comparison, the effect of estradiol (E2) and dihydrotestosterone (DHT) constantly released by Silastic implants as well as medroxyprogesterone acetate (MPA) released from poly(lactide-co-glycolide) microspheres. Femoral BMD was decreased by 11% 1 year after OVX, but treatment of OVX animals with DHEA increased BMD to a value 8% above that of intact animals. The administration of FLU reversed by 76% the stimulatory effect of DHEA on femoral BMD and completely prevented the stimulatory effect of DHEA on total body and lumbar spine BMD. Similar results were obtained for BMC. On the other hand, treatment with the antiestrogen EM-800 did not reduce the action of DHEA on BMD or BMC. At the doses used, MPA, E2 and DHT increased femoral BMD, but to a lesser degree than observed with DHEA. Bone histomorphometry measurements were also performed. While DHEA treatment partially reversed the marked inhibitory effect of OVX on the tibial trabecular bone volume, the administration of FLU inhibited by 51% (P < 0.01) the stimulatory effect of DHEA on this parameter. The addition of EM-800 to DHEA, on the other hand, increased trabecular bone volume to a value similar to that of intact controls. DHEA administration markedly increased trabecular number while causing a marked decrease in the intertrabecular area. The above stimulatory effect of DHEA on trabecular number was reversed by 54% (P < 0.01) by the administration of FLU, which also reversed by 29% the decrease in intertrabecular area caused by DHEA administration. On the other hand, the addition of EM-800, while further decreasing the intertrabecular space achieved by DHEA treatment, also led to a further increase in trabecular number to a value not significantly different from that of intact control animals, suggesting an additional effect of EM-800 over that achieved by DHEA. Treatment with DHEA caused a 4-fold stimulation of serum alkaline phosphatase, a marker of bone formation, while the urinary excretion of hydroxyproline, a marker of bone resorption, was decreased by DHEA treatment. Treatment with DHEA and DHEA + EM-800 decreased serum cholesterol levels by 22 and 65% respectively, while the other treatments had no significant effect on this parameter. The present data indicate that the potent stimulatory effect of DHEA on bone in the rat is mainly due to the local formation of androgens in bone cells and their intracrine action in osteoblasts.