Nine-month-old female rats were subjected to right hindlimb immobilization or served as controls for 0, 2, 10, 18, and 26 weeks. They were double-labeled with bone markers prior to sacrifice. Experimental unloading was produced by immobilizing the right limb against the abdomen with an elastic bandage. Single-photon absorptiometry was performed on the intact femurs; static and dynamic histomorphometry were performed on 20-micron thick toluidine blue-stained, undecalcified cross sections of the tibial shafts. Changes in the continuously immobilized tibiae were compared to those in both tibiae of age-matched controls. Unloading shut off nearly all periosteal bone formation and accelerates bone marrow expansion over that which occurs in age-related controls. The effect of unloading appeared to be mediated by recruiting fewer osteoblasts which showed inhibited activity. Furthermore, unloading increased endocortical percentage eroded surface. These histological changes lowered cortical bone mass by inhibiting diaphyseal cross sectional expansion and enlarging the bone marrow cavity. The results support Frost's suggestion that decrease mechanical usage depresses bone modeling-dependent bone gain by decreasing activation of modeling in the formation mode. It also stimulates bone remodeling-dependent bone loss by increasing activation of remodeling in the resorption mode.