Abstract: Objective To evaluate the clinical outcomes of PEEK (Polyetheretherketone) cage for surgical treatment of thoracolumbar kyphosis deformity in ankylosing spondylitis patients with andersson fracture. Methods Clinical data about 27 patients who were diagnosed with ankylosing spondylitis thoracolumbar kyphosis with andersson fracture and treated by posterior osteotomy using PEEK cage in our hospital from January 2013 to January 2015 were retrospectively analyzed. All the patients were followed up for 24-28 (26.25±1.45) months. Preoperative and postoperative radiographic data were recorded. The level height in osteotomy region (LH) and sagittal Cobb angle in globe kyphosis (GK) of the whole spine were measured on X-ray image and body fusion was evaluated. Scoliosis Research Society-22 (SRS-22) was obtained to evaluate postoperative quality of life improvement. Results All surgeries were performed successfully without severe complications. There was no significant difference between preoperative and postoperative LH (all P > 0.05). Compared with before operation, the sagittal Cobb angle in globe kyphosis of the whole spine (GK) improved significantly after operation (P < 0.05), while no significant difference was found among each time point after operation (all P > 0.05). The fusion rate was 77.8% at 3 months after operation, 96.3% at 6 months and 100% at 1 year. The SRS-22 score increased significantly at 6 months after operation (P < 0.05). No cage displacement or collapse was found during follow-up. Conclusion Posterior osteotomy through fracture line with PEEK cage is an effective method for treatment of thoracolumbar kyphotic deformity in ankylosing spondylitis patients with andersson fracture and it can avoid potential spinal cord kinking or curving. PEEK cage maintains segmental height and restoration of lordosis angle effectively during long-term follow-up with high fusion rate.