Autograft is frequently used to restore anatomic morphology and functional properties in bone defects. Disadvantages of the autograft are related to donor site morbidity and include the risk of wound infection, increased blood loss and additional postoperative discomfort. Allograft and xenograft, which are currently employed as the most common alternative to autografts, encounter the complications such as fracture, resorption and nonunion secondary to immunologic rejection. These volumetric and immunologic concerns with biologic implants have stimulated interest in the potential for synthetic, bioinert materials as bone graft substitutes. Hydroxyapatite (HA), a calcium phosphate ceramic, is a well known biocompatible artificial bone substitite without induction of systemic toxic and foreign body reactions. Bone conduction usually occurrs by the implanted HA but biodegradation of HA is poor and the bone formation around HA is slow. Carbonate apatite has been known as that the physicochemical properties are similar of the natural bony apatite and demonstrates no toxic reactions with better biodegradation. Carbonate apatite collagen composite was reported to show more bone formation and biodegradation than hydroxyap atite. In this study, the composite (CA-C) consisted of carbonate apatite and type I collagen was implanted in rabbit tibiae to evaluate the possibility as an artificial bone substitute. Forty HA (HA group) and forty CA-C (CA-C group) were applied in 80 dissected rabbit tibiae and fixed by external fixators. For biomechanical study, the rabbits were sacrificed and the specimens were obtained in 2, 4, 6 and 8 weeks after implantation. Tensile load was applied to the prepared tibiae in Instron and biomechanical properties were investigated. The fracture at the callus occurred as transverse or short oblique fracture in the vertical direction to the axis of applied tensile load. In each group, the tensile strength at breaking point increased significantly with time and at 4 weeks more prominent increase in break strength was observed (p<0.05). HA group showed higher mean strength before 6 weeks and at 8 weeks CA-C group higher mean strength but statistical significance could not be found. There was no significant difference in extension length at breaking point between both groups and time intervals. In summary, carbonate apatite collagen composite revealed similar biomechanical properties as hydroxyapatite, suggesting its clinical usefulness as a bone substitute, but it will be necessary to improve biodegradational property, stiffness of carbonate apatite collagen composite.