A soluble protein component of bone, bone morphogenetic protein, and decalcified bone matrix have been shown to induce the formation of bone in extraosseous tissue. Clinical and animal studies investigating the use of these materials as bone grafts have shown radiographic and histological evidence of formation of bone, but the clinical usefulness of these grafts remains unknown. This study compared the healing processes when plasma-coated demineralized bone matrix and autologous cancellous bone were used to graft segmental defects of bone. A standard procedure was used to make a two-centimeter defect bilaterally in the ulna of forty-eight skeletally mature New Zealand White rabbits. In each rabbit, one ulnar defect was grafted with autologous citrated plasma-coated demineralized bone matrix while the other defect served as a control and was grafted with either autologous cancellous bone from the iliac crest, demineralized bone matrix, or demineralized bone matrix augmented with bone proteins that had been extracted with guanidinium hydrochloride. The ulnar defect was stabilized by the intact radius, and no supplemental device was necessary for fixation. To examine spontaneous healing in this model, one group of rabbits had a control defect that was not grafted. The grafts were periodically evaluated by radiographs, and twelve weeks after surgery the grafts were harvested and tested to failure in a standard torsion-test machine. The mechanical parameters were calculated, and histological examination of major fragments of the grafts was performed. The results of the radiographic and histological evaluation showed that all of the grafted ulnae healed, with fusion of the graft to the cut ends of the defect and reformation of approximately normal anatomy. No ungrafted ulnar defects healed. The results from the mechanical tests were evaluated by comparing the defect that was grafted with plasma-coated demineralized bone matrix with the control graft in each animal. These data showed that: twelve weeks after grafting, the normal ulnae were significantly stronger than the ulnae that had been grafted with plasma-coated demineralized bone matrix; the ulnae that had been grafted with plasma-coated demineralized bone matrix and those that had been grafted with autologous bone were equivalent in strength; and twelve weeks after grafting, grafts of demineralized bone matrix that were augmented with extracted bone proteins were significantly stronger than those that had not been so augmented.