The relative clinical efficacy of surface-decalcified and wholly decalcified bone alloimplants was studied by observing radiographically the incorporation of the two types into the recipient bone bed in cases of shelf arthroplasty of the hip joint. In all cases, the decalcified alloimplants were incorporated well in an intraskeletal site. However, in all five cases examined, wholly decalcified alloimplants in a paraskeletal site were resorbed with little or no new bone deposition. In contrast, surface-decalcified alloimplants in a paraskeletal site were consistently incorporated, though their complete integration required more than one year. These findings show that, contrary to the results in experimental animals, increased decalcification of bone alloimplants reduced their clinical efficacy.