The objective of this study was to evaluate the long term outcome of combined anterior and posterior convex spinal epiphysiodesis in the treatment of congenital scoliosis. The study covered 53 patients (27 male, 26 female) with a minimum follow up period from surgery of 3 years (mean 8.8 years, range 3-22.5 years). Of these, 34 were skeletally mature when reviewed. Clinical assessment and sequential measurement of Cobb angles were used to chart the course of the deformity following surgery. The types of vertebral anomalies encountered were: 4 unsegmented bars, 7 unsegmented bars with hemivertebrae, 30 hemivertebrae (of which 2 were double hemivertebrae) and 12 complex (unclassifiable) patterns. The severe types were concentrated in the thoracic spine. Results are presented with reference to the type and site of anomaly and to the age of the patient at the time of surgery. Where deformity was due to an unsegmented bar (with or without hemivertebra) the rate of change of Cobb angle was slowed, but not reversed, following surgery. For the complex anomalies there was a reduction in the rate of progression of deformity following surgery, however, the final Cobb angle still increased from a mean of 61 degrees to 70 degrees. In contrast, the rate of progression reversed or slowed in 97% of the hemivertebra patients following surgery, producing a change in mean Cobb angle from 41 degrees pre-operatively to 35 degrees post-operatively. For each type of anomaly the correction achieved was greater where the surgery was performed at a younger age. Final outcome was influenced by the site of anomaly, with a better correction being achieved in the lumbar than the thoracic spine. In conclusion, we feel that convex epiphysiodesis has an important role in the surgical management of congenital scoliosis and, for hemivertebrae in particular, it can produce significant correction of deformity.