METHODS Retrospective review of patient records with current clinical and radiographic assessment. OBJECTIVE To evaluate the long-term result of anterior and posterior convex spinal growth arrest, with or without instrumentation, in managing infantile idiopathic scoliosis. BACKGROUND There were 12 male and 10 female patients studied, with a mean follow-up period of 10 years, 9 months. The mean Cobb angle before surgery was 65 degrees. All had a rib vertebral angle difference more than 20 degrees. The mean age at surgery was 6 years. Nine patients had epiphysiodesis alone; nine patients also underwent Harrington instrumentation simultaneously, and four underwent Harrington instrumentation 2-4 years later. METHODS Clinical evaluation and sequential measurements of Cobb angle were done. RESULTS The epiphysiodesis-only group had a mean preoperative Cobb angle of 72 degrees, mean progression of curves of +12 degrees, and mean rate of progression of +2.5 degrees per year: the group's postoperative figures were 92 degrees, +15 degrees, and +3 degrees per year, respectively. The epiphysiodesis and late Harrington rod group had a mean preoperative Cobb angle of 56 degrees, mean progression of +12 degrees, and a mean rate of progression of +5 degrees per year; the group's postoperative Cobb angle averaged 62 degrees, progression +6 degrees, and rate of progression +1 degree per year. The epiphysiodesis with simultaneous Harrington rod group had a preoperative mean Cobb angle of 60 degrees, mean progression of +18 degrees, and mean rate of progression of +6 degrees per years. After surgery, these improved to 58 degrees, correction of 2 degrees, and rate of correction of 0.5 degree per year. CONCLUSIONS Combined anterior and posterior convex spinal growth arrest alone does not prevent progression of deformity in infantile idiopathic scoliosis. The addition of posterior instrumentation can slow or arrest deformity progression but not reverse it.