A prospective cohort study of consecutive patients with thoracolumbar fractures from two centers was conducted to evaluate the use of cortical allograft for anterior spine reconstruction. The purpose was to determine whether allograft bone had deleterious effects on fusion rates and maintenance of correction. Thirty-six patients were followed for a minimum of 2 years. A radiologic fusion rate of 81% was found, and there was a loss of postoperative correction in eight patients. The addition of posterior instrumentation was not significantly different from anterior surgery alone with respect to union rates but did provide a more reliable maintenance of correction of deformity. There were no cases of disease transmission. The authors conclude that cortical allograft for anterior spine reconstruction in trauma is safe and despite having a fusion rate lower than autogenous iliac crest, little clinical significance is attributable to the radiologic appearance of nonunion at the host-allograft junctions.