BACKGROUND: Repair of bone defects is not only a clinical problem, but also a hot topic in the field of orthopedics. Although autologous bone grafting is considered as the "gold standard" for bone repair, its use is limited due to the limited source of autogenous bone, bone infections and pains that are easy to occur in the donor region. Allograft bones are always associated with immune rejection, slow healing, and infection. Therefore, it is imminent to develop new materials for bone repair. OBJECTIVE: To explore the effect of rabbit adipose-derived stem cells (rADSCs) as seed cells and hydroxyapatite/β-tricalcium phosphate (HA/β-TCP) composite as a carrier on the repair of rabbit vertebral defects. METHODS:Thirty-eight 3-month-old New Zealand white rabbits were selected,and two of them were used to culture rADSCs in vitro. Passage 3 rADSCc were inoculated on HA/β-TCP scaffolds and then cultured in vitro for 2 weeks.A 5 mm×5 mm×3 mm bone defect was prepared at the anterior edge of L4/5vertebral body in the remaining 36 rabbits. These model rabbits were then randomized into cell-scaffold composite group, scaffold group and control group with no intervention, with 12 rabbits in each group. rADSCs/HA/β-TCP composite and HA/β-TCP scaffold were implanted into the cell-scaffold and HA/β-TCP groups, respectively. Anteroposterior and lateral DR of the spine and Lane-Sandhu X-ray were performed at 4, 8, 12 postoperative weeks. All rabbits were sacrificed at 12 postoperative weeks and specimens were collected for gross and histopathological observations. RESULTS AND CONCLUSION: Under the gross observation, bone defects in the cell-scaffold group were essentially replaced by new bone tissues, which was significantly better than that in the scaffold group and control group. At 12 postoperative weeks, the material implanted was basically absorbed in the cell-scaffold group, partially absorbed in the scaffold group and poorly absorbed in the control group in which there was a clear boundary with the surrounding tissues and patchy calcified shadows were visible. X-ray results showed that the repair effect in the cell-scaffold group was better than that in the scaffold group and control group (P < 0.05). Histopathological findings showed the marked absorption of the implant in the cell-scaffold group, partial residual in the scaffold group with some fibrous calluses and osteoid tissues, and a large amount of fiber tissues and a small amount of calluses in the control group. Overall, the rADSCs/HA/β-TCA has a good ability to repair bone defects.