An increase in osteoblast and osteocyte apoptosis has been demonstrated in mice and humans receiving glucocorticoids and may be involved in the pathogenesis of the associated osteonecrosis. To examine the spatial relationship between osteocyte apoptosis and glucocorticoid-induced osteonecrosis, we determined the prevalence of osteocyte apoptosis in whole femoral heads obtained from patients who underwent prosthetic hip replacement because of osteonecrosis due to chronic glucocorticoid treatment (n = 5), alcoholism (n = 3), and trauma (n = 1) as well as in femoral neck cores from patients with sickle cell disease (n = 5). Abundant apoptotic osteocytes and cells lining cancellous bone were found juxtaposed to the subchondral fracture crescent in femurs from the patients with glucocorticoid excess. In contrast, apoptotic bone cells were absent from the specimens taken from patients with trauma or sickle cell disease and were rare with alcohol abuse. These results indicate that glucocorticoid-induced osteonecrosis is a misnomer. The bone is not necrotic; instead, it shows prominent apoptosis of cancellous lining cells and osteocytes. Glucocorticoid-induced osteocyte apoptosis, a cumulative and irreparable defect, could uniquely disrupt the mechanosensory function of the osteocyte network and thus start the inexorable sequence of events leading to collapse of the femoral head.