PURPOSE: To determine optimal levels of posterior fixation in thoraco-lumbar bursting fractures according to the Load-sharing classification. MATERIALS AND METHODS: From Aug. 1999 to Aug. 2003, 50 patients who had been operated with the posterior fixation in one-body thoraco-lumbar bursting fracture were selected. They were divided into two groups, group I, 6 points and below in the Load-sharing score and group II, 7 points and above. And also, each groups subdivided into two subgroups, A (short segment fixation including below and above one body) and B (long segment fixation including below and upper two body). So patients subdivided into I-A, I-B, II-A, II-B. Change of the corrected kyphotic angle was measured and compared with each subgroups. RESULTS: The loss of the corrected kyphotic angle was measured average 1.7degrees in group I and 4.1degrees in group II, and there was significant difference between two groups (p>0.05). The loss of the corrected kyphotic angle in the subgroups was average 1.8degrees in I-A, 1.6degrees in I-B, 3.5degrees in II-A and 4.9degrees in II-B. And there was significant difference statistically in I-A and II-A (p>0.05). CONCLUSION: In the thoraco-lumbar bursting fracture with 6 points and below of the Load-sharing score, the fixation of the short segment is a useful method. But in the fracture with 7 points and above, the fixation of the short segment is not enough, and these findings be required the further evaluation for some cause of the loss of corrected angle and treatment modalities including the fixation of the long segment.