W e used a model whereby mechanical stimulation induces bone formation in rat caudal vertebrae, to test the effect of estrogen on this osteogenic response. Unexpectedly, estrogen administered daily throughout the experiments (8-11 d) suppressed, and ovariectomy enhanced, mechanically induced osteogenesis. Osteogenesis was unaffected by the resorption-inhibitor pamidronate, suggesting that the suppression of bone formation caused by estrogen was not due to suppression of resorption. We found that estrogen did not significantly reduce the proportion of osteocytes that were induced by mechanical stimulation to express c-fos and IGF-I mRNA; and estrogen suppressed mechanically induced osteogenesis whether administration was started 24 h before or 24 h after loading. This suggests that estrogen acts primarily not on the strain-sensing mechanism itself, but on the osteogenic response to signals generated by strain-sensitive cells. We also found that when estrogen administration was started 3 d after mechanical stimulation, by which time osteogenesis is established, estrogen augmented the osteogenic response. This data is consistent with in vitro evidence for estrogen responsiveness in two phenotypically distinct bone cell types: stromal cells, whose functional activities are suppressed, and osteoblasts, which are stimulated, by estrogen.