STUDY DESIGN: A randomized, controlled follow-up study. OBJECTIVE: The objective of this study was to compare the results of anterior approach versus posterior approach with subtotal corpectomy, decompression, and reconstruction of spine in the treatment of thoracolumbar burst fractures. SUMMARY OF BACKGROUND DATA: Burst fractures are frequently associated with instability or neurological deficit. Anterior subtotal corpectomy, decompression, and reconstruction with instrumentation are an established method for a highly unstable burst fracture. In the past few years, subtotal corpectomy, decompression, and reconstruction of spine could be completed by posterior approach. Posterior segmental pedicle screw instrumentation, with its more rigid fixation and less technically demanding, could offer potential advantages. METHODS: A total of 64 patients with thoracolumbar burst fractures were divided into 2 groups randomly. Group A was treated by anterior approach and group B was treated by posterior approach with subtotal corpectomy, decompression, and reconstruction of spine. During the minimum 24 months (range, 24 to 72 mo) follow-up period, all patients were prospectively evaluated for clinical and radiologic outcomes. The intraoperative blood loss, operative time, complications of operation, pulmonary function, Frankel scale, and the American Spinal Injury Association (ASIA) motor score were used for clinical evaluation, whereas the heights of anterior edge of vertebral body and the Cobb angle were examined for radiologic outcome. RESULTS: All patients in this study achieved solid fusion, with significant neurological improvement. The intraoperative blood loss (P<0.05) and complications of operation were less, the operative time was shorter (P<0.05), and the pulmonary function after operation was better in the group B (P<0.05). The Frankel scale, the ASIA motor score, and the radiologic results were not significantly different (P<0.05) at all time points between the 2 groups A and B. But the 2 groups improved in their neurological function by approximately 1.3 Frankel grade and 15.6 ASIA motor scores at final follow-up. CONCLUSION: Anterior approach and posterior approach with subtotal corpectomy, decompression, and reconstruction of spine are sufficient for surgical treatment of thoracolumbar burst fractures. Less intraoperative blood loss and complications, shorter operative time, and better pulmonary function after operation are the significant advantages of posterior surgery.