Over a 2 1/2-year period, I have treated 25 patients with segmental spinal instrumentation (SSI), using paired-wired Luque rods. Of 13 patients with unstable spinal fractures, seven were paraplegic. Nine had idiopathic scoliosis, and three had neurogenic scoliosis. Significant loss of correction after SSI occurred in five of the nine patients with idiopathic scoliosis and in two of the three with neurogenic scoliosis. In vitro testing of Harrington distraction rods and paired-wired L-shaped rods showed the Harrington rod to resist shortening and lateral bending due to axial compression forces better than the Luque system. From the study I conclude that postoperative brace protection is as necessary to the success of SSI as with the more traditional Harrington technique. Segmental spinal instrumentation is a powerful corrective device, but appears to be deficient in maintaining correction.