STUDY DESIGN: A prospective, randomized study comparing two treatment methods for thoracolumbar burst fractures: short-segment instrumentation with transpedicular grafting and the same procedure without transpedicular grafting. OBJECTIVE: To evaluate the efficacy of transpedicular grafting in preventing failure of short-segment fixation for the treatment of thoracolumbar burst fractures. SUMMARY OF BACKGROUND DATA: Short-segment pedicle instrumentation for thoracolumbar burst fractures is known to fail early because of the absence of anterior support. Additional transpedicular grafting has been offered as an alternative to prevent this failure. However, there is controversy about the results of transpedicular grafting. METHODS: Twenty patients with thoracolumbar burst fractures were included in the study. The inclusion criterion was the presence of fractures through the T11-L3 vertebrae without neurologic compromise. The patients were randomized by a simple method into two groups. Group 1 patients were treated using short-segment instrumentation with transpedicular grafting (TPG) (n = 10), and Group 2 patients were treated by short-segment fixation alone (NTPG) (n = 10). Clinical (Likert's questionnaire) and radiologic (sagittal index, percentage of anterior body height compression, and local kyphosis) outcomes were analyzed. RESULTS: The two groups were similar in age, follow-up period, and severity of the deformity and fracture. The postoperative and follow-up sagittal index, percentage of anterior body height compression, and average correction loss in local kyphosis in both groups were not significantly different. The failure rate, defined as an increase of 10 degrees or more in local kyphosis and/or screw breakage, was also not significantly different (TPG = 50%, NTPG = 40%, P = 0.99). CONCLUSIONS: Short-segment transpedicular instrumentation of thoracolumbar burst fractures is associated with a high rate of failure that cannot be decreased by additional transpedicular intracorporeal grafting.