In adolescent idiopathic scoliosis the cosmetically unacceptable rib prominence is one of the main reasons patients seek treatment. Ninety-eight patients were reviewed with Cotrel-Dubousset (CD) instrumentation (average follow-up, 27 months; average preoperative curve, 53 degrees) and 27 patients with Harrington instrumentation and rib resection (average follow-up, 43 months; average preoperative curve, 52 degrees). Of the 98 patients who had CD instrumentation, 15 underwent concomitant rib resection. Of the 83 patients without rib resection, 71 (72%) were rated satisfactory and 12 unsatisfactory by cosmetic criteria based on residual rib deformity. These results were compared to those of 27 patients receiving Harrington rod instrumentation and rib resection, of whom 23 (85%) were rated as satisfactory. All 15 patients with CD and rib resection were rated satisfactory. The CD patients were then redivided into two groups (rib resection indicated or rib resection not indicated) as follows: the 12 unsatisfactory CD patients without rib resection (in whom a rib resection should have been done) were grouped with the 15 CD patients who underwent rib resection, for a total of 27 rib resections indicated, or 28%. This group was compared to the 71 satisfactory CD patients without rib resection (rib resection not indicated). Patients with a rib prominence of > 15 degrees preoperatively had or should have had a rib resection. Patients with a higher chance of needing rib resection included those with a curve severity greater than 60 degrees, curve flexibility less than 20%, a preoperative rib prominence > 10 degrees, or intraoperative curve correction of less than 50%.