Objectives: To investigate the long-term efficacy of balloon assisted endplate reduction with vertebral augmentation combined with pedicle screw fixation in the treatment of thoracolumbar burst fractures, and to compare the clinical efficacy of calcium sulfate cement (CSC) and calcium phosphate cement(CPC). Methods: This study is a retrospective cohort study.The clinical data of 39 patients with thoracolumbar burst fractures admitted to Hefei Hospital Affiliated to Anhui Medical University from November 2013 to December 2017 were retrospectively analyzed.All patients were treated with pedicle screw reduction and fixation of the injured vertebra,balloon-assisted reduction of the collapsed endplate of the injured vertebra,and artificial bone vertebral body augmentation,and the follow-up time was >5 years.There were 24 males and 15 females,aged (42.9±13.3) years (range: 29 to 56 years).According to the Frankel spinal nerve dysfunction grading standard, there were 4 cases of grade C, 7 cases of grade D and 28 cases of grade E. There were 21 cases of CSC augmentation(CSC group) and 18 cases of CPC augmentation (CPC group). X-ray and CT were performed at 1 week, 1-, 2-, 5-year after surgery and at the last follow-up, and the imaging indicators were measured, including the injured vertebra anterior edge height ratio,the injured vertebra middle height ratio,the injured vertebra wedge angle,and the sagittal plane Cobb angle. The pain visual analogue scale (VAS) and the Oswestry disability index (ODI) was used for functional evaluation, nervous function was evaluated according to the Frankel spinal nerve dysfunction grading standard.Independent sample t test was used for inter-group comparison, and paired sample t test and repeated measure ANOVA were used for intra-group comparison. Results: All operative procedures were successfully completed, no spinal nerve function damage occurred. The postoperative imaging indexes of the patients were significantly improved compared with those before surgery (all P<0.01). The follow-up time of patients was (6.7±2.8)years (range: 5 to 9 years). Among the 11 patients with symptoms of neurological impairment before surgery, 9 patients completely recovered at the last follow-up, and 2 patients recovered from Frankel grade C to D. There were no significant differences in imaging indexes between the first week after surgery and the last follow-up in the CPC group (all P>0.05), while there were significant differences in imaging indexes between the CSC group and the last follow-up (all P<0.05). CPC group was superior to CSC group in frontal height ratio, middle height ratio, wedge angle variation and sagittal Cobb angle correction loss at 2 year, 5 year after surgery and the last follow-up, with statistical significance (all P<0.05). At the last follow-up, there were no differences in VAS and ODI between the two groups (all P>0.05). After absorption of CSC in the filling area, a hardened zone was formed around the area, and the central cavity remained without bone tissue filling. CPC absorption was very slow, and the CPC group was still filled satisfactorily at the last follow-up. Conclusions: Balloon assisted endplate reduction and vertebral augmentation combined with pedicle screw fixation through the injured vertebra have good long-term efficacy in the treatment of thoracolumbar burst fractures. Compared with CSC, CPC vertebral augmentation can better maintain the shape and spinal sequence of the injured vertebra in the long term, and can effectively reduce the collapse of the space above the injured vertebra.