We studied whether the pedicle screw is better than laminar hooks for fixation of the lumbar spine in the treatment of idiopathic scoliosis. 66 consecutive patients with idiopathic scoliosis (King I and II) were studied retrospectively. Group S included 33 patients (25 females) treated with pedicle screws. Their mean age at operation was 17 (13-54) years. Group H included 33 patients (30 females) treated exclusively with hooks. Their mean age at operation was 16 (1140) years. The preoperative mean angles of the thoracic curve in group S was 66 (42.115) degrees, and in group H 65 (42-121) degrees. The lumbar curve averaged 46 (20-85) degrees in group H and 53 (33-86) degrees in group S. All patients were fused only posteriorly with Cotrel-Dubousset instrumentation and an autogenic bone graft. The mean follow-up time was 4 (2-7) years. Mean correction of the thoracic curve was 45% in group S and 50% in group H. The lumbar curve was corrected by 50% in group S and 51% in group H. Loss of correction of the thoracic curve occurred in 5% in group S and 6% in group H and of the lumbar curve in 3% in group S and 10% in group H (p = 0.04). Group S better maintained the correction of the lateral tilt of the uninstrumented segment adjacent to the fusion (p = 0.04). Derotation, according to Perdriolle, in the distal segment adjacent to the fusion was 6% in group S and 2% in group H. We found no difference between correction of the thoracic and lumbar curves using pedicle screws and laminar hooks in the lumbar spine. Pedicle screws better maintained the correction of the lumbar curve and the lateral tilt in the distal segment adjacent to fusion.