Of 229 patients, 15 pairs of patients with right thoracic idiopathic sco-liosis (King III) and 15 pairs with combined thoracic and lumbar scoliosis (King II) were identified. Each pair consisted of one patient with Harrington and the other with Cotrel-Dubousset instrumentation. Pairs were comparable with regard to thoracic curve magnitude, and level of end and stable vertebrae. Minimum follow-up was 4 years postoperatively. Spinal mobility of patients in the coronal and sagittal plane, and axial rotation was measured with inclinometers. The differences between operatively treated groups are not significant. The more caudal instrumentation ends, the more reduced is spinal mobility in the coronal plane. Motion in operated patients is significantly reduced in comparison to a healthy reference group.