Objective To review the King-types Ⅲ and Ⅳ patients treated by the CD hybrid technique and evaluate clinical results on the shorter fusion levels. Methods Fifty-eight patients with idiopathic scoliosis were treated by the CD hybrid method from March 2000 to January 2003, among whom 40 were grouped as Kingtype Ⅲ and 18 as Kingtype Ⅳ; 41 were female and 17 were male. The Cobb angle of the thoracic curve was averaged 64°(range 50-83°), and the curve flexibility was 62%. The compensative lumbarcurve was averaged 37°（range 16-48°）, and the curve flexibility was 105%. With the neutral rotational vertebrae as a basis to select the low instrumentation vertebrae, the neutral rotational vertebrae or the vertebrae at 1 or 2 levelsproximal to the neutral rotational vertebrae were selected as the low instrumentation vertebrae in all the patients. Standing AP and lateral radiographs were taken respectively at the discharge, during the follow-up after discharge, and at the final follow-up. Results The patients were followed up for an average of 2.4 years (range 1.8-3.2). The corrected curves lost an average of 3.1°（range -1-5°）and the correction rate of the thoracic curve was 68% at the final follow-up.The plumbline from C7 was parallel to the sacral midline in 56 patients. The lumbar curves were corrected to an average of 8°（2-13°）automatically. The lumbosacral angle was corrected automatically by 53% and the low instrumentation vertebrae in 48 patients turned into stable vertebrae. The low instrumentation vertebrae lost 1.4 segments on average compared with the Harrington principle. No spinal imbalance was clinically observed in all the patients. Conclusion The choice of the low instrumentation vertebrae as the neutral rotational vertebrae can have a good result in the clinical practice. It can be applied in the CD hybrid technique in treatment of idiopathicthoracic curves.