Objective: To evaluate the biomechanical performance of vertebroplasty using calcium sulfate cement for thoracolumbar burst fractures.Methods: Sixteen bovine thoracolumbar spines (T11-L1 ) were divided into 4 groups (A, B, C and D). After burst-fracture model was created, 12 vertebral bodies in Groups A, B and C were augmented with calcium sulfate cement (CSC), calcium phosphate cement (CPC) and polymethylmethacrylate ( PMMA ) bone cement,respectively. Each anterior vertebral body height was measured with a caliper at 4 time points: intact conditions(HInt), post-fracture (HFr), post-reduction (HRe) and post-vertebroplasty (HVP). The filling volume of 3 different bone cements was also measured. Each vertebral body was compressed at 0.5 mm/s using a hinged plating system on a materials testing machine to 50% of the postvertebroplasty height to determine strength and stiffness.Difference was checked using t test or One-way ANOVA.Results: The average strike energy was 66. 2 J.Vertebroplasty with different cements could sustain vertebral height. The average filling volume of bone cement in 3 groups was 4.35 ml (CSC), 3.72 ml (CPC) and 3.95 ml (PMMA), respectively, and there was no statistically significant difference among them ( P ＞0.05).Vertebroplasty with PMMA completely restored strength(116%) and stiffness (105%). CSC or CPC partly recovered vertebral strength and stiffness. However,greater strength restoration was got with CSC ( 1 659 N) as compared with CPC (1 011N, P＜0.01 ). Regarding stiffness, differences between CSC (140 N/mm ±40 N/mm)and the other two bone cements ( CPC :148 N/mm ±33 N/mm,PMMA:236 N/mm ±97 N/mm) were not significant (P＞0.05).Conclusions: For a burst-fracture of calf spine, useof CSC for vertebroplasty yields similar vertebral stiffnessas compared with PMMA or CPC. Although augmentationwith CSC partly obtains the normal strength, thistreatment still can be applied in thoracolumbar burstfractures with other instrumental devices in light of itsbioactivation.