Demineralized bone powder was used as an osteoinductive substance to bridge very large defects (more than 50 per cent of the total length of the bone) in one radius of each of thirty-three rats. An identical defect was produced in the contralateral radius of each animal for use as a control. The defect on the control side was left unbridged or was bridged by large chips of autologous bone or an autologous inlay graft. All rats showed formation of new bone throughout the length of the radial defect only on the side in which the demineralized bone powder had been implanted. The control side, in which an autologous graft in the form of chips or inlay had been implanted, showed resorption of the graft. The maximum rate of formation of bone occurred fifteen to twenty-one days after implantation of the demineralized bone powder. At thirty-five days, the experimental defect was fully bridged, forming solid bone, in 71 per cent of the rats, and the remaining 29 per cent showed bridging of 95.8 per cent of the length of the defect, with union on one side. Analysis of the sequential radiographs, technetium-99m scans, and histological findings showed that the formation of bone and bridging of the defect were superior on the side in which the demineralized bone powder had been implanted compared with the side in which pieces of autologous bone or an autologous inlay graft had been used.(ABSTRACT TRUNCATED AT 250 WORDS)