This study was a 15-20 year follow-up of a previous trial that compared operative and nonoperative treatment for thoracolumbar burst fractures. From 1992 to 1997, 47 patients were allocated to either operative (posterior or anterior arthrodesis) or nonoperative (body cast or orthosis) treatment. Early analysis (4 years) from the previous trial revealed few significant differences between the groups. However, the late follow-up (15-20 years) analysis indicated that the nonoperative group reported less pain and better function compared to those treated surgically for thoracolumbar burst fractures.