METHODS Retrospective radiographic review. OBJECTIVE To determine whether clinical factors or common classification systems can predict the radiologic outcome of short-segment thoracolumbar fracture fixation. BACKGROUND Previous reports have indicated that short-segment thoracolumbar fracture fixation might not be appropriate for highly comminuted fractures or for patients with multiple traumatic injuries. METHODS We conducted a retrospective radiographic review of 46 thoracolumbar fractures treated with short-segment posterior instrumentation to determine the rate of correction loss and instrumentation failure in relation to the Load Sharing Classification of Spine Fracture system and the AO Classification of Fractures system. No postoperative bracing was used. Patients with multisystem organ trauma and those with isolated injuries were included. RESULTS An average loss of correction of 7.5 degree was observed. Pedicle screw placement into the fractured vertebra seemed to protect against correction loss: 4 of the 7 patients (57%) in the no intermediate fixation group had >10 degree loss of correction. No relationship was shown between loss of correction and Load Sharing Classification (< or = 6 or > or = 7 points), loss of correction and AO Classification, or loss of correction and level of injury (thoracolumbar junction vs. lower lumbar). CONCLUSIONS With modern instrumentation and techniques, short-segment thoracolumbar fracture fixation could be used successfully, despite highly comminuted injuries, without anterior column support or supplemental bracing.