Objective To evaluate the osteogenic effect of dextran sulfate/recombinant human bone morphogenetic protein 2/chitosan (DS/rhBMP-2/CS) combined with coralline hydroxyapatite (CHA) in repairing large segmental bone defects by radiographic feature. Methods Fifty-seven 24-week-old male New Zealand rabbits were prepared for establishing right radius bone defect model of 20 mm in length. In which 54 rabbits were randomly divided into 3 groups (n=18), and the CHA, DS/rhBMP-2/CS/CHA, and rhBMP-2/CHA artificial bone grafts were implanted into the bone defect in groups A, B, and C respectively; the remaining 3 rabbits were implanted nothing as blank control group. After operation, the gross condition of the animals was observed; at 4, 8, and 12 weeks after operation, X-ray film observation, Micro-CT scanning, and three-dimensional reconstruction were performed to obtain the volume of the new bone. Results The experimental animals recovered well and were in normal condition. X-ray observation showed that the bone healing in group B was better than that in groups A and C at each time point. At each time point after operation, the X-ray scores of group B were significantly higher than that of group A and group C (P<0.05); the scores of group C at 8 and 12 weeks after operation were significantly higher than that of group B (P<0.05). Micro-CT scanning and three-dimensional reconstruction observation showed that at each time point after operation in group A, the bone defect area had less bone formation and poor osteogenesis; in group B, there were many new bone tissues in bone defect area, and the bone remodeling was well, and gradually closed to normal bone morphology at 12 weeks; in group C, there were many new bone tissues in bone defect area, but the bone formation was general. The new bone volume of group B was significantly higher than that of group A and group C (P<0.05) at each time point after operation, and the score of group C was higher than that of group A at 8 weeks after operation (P<0.05). Conclusion The osteogenic effect of DS/rhBMP-2/CS/CHA sustained-release artificial bone is much better than that of single CHA and rhBMP-2/CHA, which can provide a new idea for treating bone defect by using bone tissue engineering in the future.