OBJECTIVE To study the difference of repairing segmental bone defect with bio-derived bone preserved by various methods. METHODS Freeze-dried biomaterials had been stored in two different preservation solutions for three months, while the biomaterials stored for same period were observed as control group. The experimental model of 15 mm radial segmental defect was made in 60 New Zealand white rabbits, which were divided into groups A, B and C according to transplant materials preserved by various methods. Groups A and B were deeply divided into A1 and A2 subgroups, B1 and B2 subgroups according to whether materials were cocultured with osteoblasts. Tissue engineered bone was used to repair bone defects of left limbs in A1 and B1 subgroups, while simple material to repair defects of right limbs in A2 and B2 subgroups. Group C was divided into C1 and C2 subgroups. Freeze-dried material was used to repair bone defects of the left limbs, while defects of the right limbs as blank control group. The samples were harvested and observed by the roentgenographical, histomorphological, biomechanical and computerized graphical analysis at 4, 8 and 16 weeks. RESULTS All of the defects treated with implants exhibited new bone formation 4, 8 and 16 weeks postoperatively, increasing with time. The radiological, histomorphological and biomechanical evaluation showed that the ability of new bone formation was arranged in 6 subgroups as follows: A1>A2>C1>B1>B2>C2, the difference was significant between them (P<0.001, P<0.05). The ability of new bone formation was strongest and at 16 weeks the defect was bridged with the appearance of marrow cavities in A1 subgroup, the biomechanical properties in implants approached to those of normal bone. CONCLUSIONS The choice of proper preservation solution can improve the ability of repairing bone defect.