STUDY DESIGN: A retrospective study. OBJECTIVES: To evaluate the relationships between spinal canal occlusion and neurologic deficits, and between spinal canal decompression and neurologic recovery in thoracolumbar burst fractures. Kyphotic deformities, based on the fracture types in short-segment instrumentation and fusion, were evaluated to determine effective operative methods. SUMMARY OF LITERATURE REVIEW: In thoracolumbar burst fractures, the relationship between spinal canal occlusion and neurologic deficits remains controversial; and definitive guidelines for short-segment instrumentation and fusion have not been established . MATERIALS AND METHODS: Surgically treated thoracolumbar burst fractures (N=112) were analyzed retrospectively. Spinal canal occlusion in both neurologically intact and deficient groups, and neurologic recovery as a result of spinal canal decompression, were evaluated based on Frankel's grades. Kyphotic deformities based on the Denis classification and McCormack's load sharing classification were evaluated in 86 short-segment instrumentation patients. RESULTS: Spinal canal occlusion in the neurologically deficient group (51.8%) was significantly higher than that in the neurologically intact group (31.4%) (p < 0.05). Although 29 patients who recovered neurologically and 25 who did not, demonstrated 20.4% and 19.5% of spinal canal decompression, respectively, it was not significant (p > 0.05). Kyphotic deformities were increased significantly in Denis type A, B and groups with more than 7 points in the load sharing classification (p < 0.05). CONCLUSION: In thoracolumbar burst fractures, the degree of initial spinal canal occlusion was more significantly related with neurologic deficits than with postoperative spinal canal decompression. Extended instrumentation and fusion is recommended for reducing postoperative kyphotic deformities in Denis type A, B and groups with more than 7 points in the load sharing classification.