The osteoconductive and osteoinductive potential of two human allogeneic demineralized bone matrix putties were compared in a critical-sized athymic rat femoral defect model. Defects were treated with (1) a demineralized bone matrix in a hyaluronic acid carrier, (2) a demineralized bone matrix in a glycerol carrier, (3) a hyaluronic acid carrier alone, or (4) with no implant. Radiographic examinations and histologic analyses were done at 4, 8, and 16 weeks postoperatively. Eight of the 48 defects treated with a demineralized bone matrix and none of the 36 surgical controls showed complete radiographic healing by 16 weeks and no statistically significant difference between the radiographic scores for the two demineralized bone matrix preparations was found. On histologic review, both preparations of demineralized bone matrix had passive remineralization. The largest foci of endochondral ossification were seen in limbs treated with a demineralized bone matrix in a hyaluronic acid carrier. The 8-mm rat femoral defect allows for stringent assessment of the osteoinductive potential of bone graft substitutes. Hyaluronic acid and glycerol are viable carriers for demineralized bone matrices. As both de-mineralized bone matrices tested provided an adequate osteoconductive matrix and showed some, although limited, osteoinductive capacity, these materials should be used in clinical practice only as bone graft extenders or enhancers.