We have known for sometime that sex hormones influence the growth, preservation, and loss of bone tissue in the skeleton. However, we are only beginning to recognize how estrogen influences the responsiveness of the skeleton to exercise. Frost's mechanostat theory proposes that estrogen reduces the mechanical strain required to initiate an osteogenic response, but this may only occur at the endocortical and trabecular bone surfaces. The discovery of estrogen receptors alpha and beta may help us to understand the bone surface-specific effects of exercise. Findings from estrogen receptor knockout mice suggest that the activity of ERalpha may explain the positive interaction between estrogen and exercise on bone formation near marrow, that is, endocortical and trabecular bone surfaces. Estrogen inhibits the anabolic exercise response at the periosteal surface, and this we propose is due to the activation of ERbeta. Signaling through this receptor retards periosteal bone formation and suppresses gains in bone size and bone strength, and for these reasons it behaves as an antimechanostat.