Segmental spinal instrumentation with Harrington rod secured to the spine by sublaminar wires was a popular method of scoliosis correction in 1980's. It was gradually replaced by newer rod-hook systems due to concern about neurological complications. However, correction of type II and III curves by selectively fusing the thoracic curves with these new instruments has resulted in poor results in some cases. The aim of this study is to review the result of selective thoracic fusion treated by segmental spinal instrumentation. Between January 1989 to October 1994, 31 patients with King II scoliosis were treated operatively in our unit. These consisted of 29 girls and 2 boys. The mean age of these patients were 11.3 years. The study population consisted of 21 Chinese, 5 Malays and 5 Indians. In one patient, the thoracic curve was convex to the left whilst the thoracic curves in the majority were to the right. The surgery was performed by three surgeons using harrington rods and posterior fusion with autograft. Anterior releases were also required in eight patients to increase flexibility. The curve correction obtained was an improvement from a average preoperative cobb's angle of 71.5 degrees to 39.5 degrees postoperatively. After an average follow-up period of 77.9 months, the correction deteriorated by 22% in the thoracic curve and 59% in the lumbar spine without disturbance to truncal balance. Only one sublaminar wire broke. However, no implant failure or removal has to be performed as yet. This technique appears useful in our institution with minimal morbidity.