To determine the risk factors for scoliosis in children treated for neuroblastoma. From 1957 to 1997, 58 children with neuroblastoma were treated at 1 institution and have survived a minimum of 5 years. There were 35 boys and 23 girls with a median age of 6 months (range, 2 weeks to 15 years) at initial diagnosis. Primary site was located in the adrenal gland in 25 (43.1%), abdominal /non-adrenal in 16 (27.6%), thoracic in 12 (20.7%), cervical in 3 (5.3%) and pelvic region in 2 (3.5%). The International Neuroblastoma Staging System (INSS) stage was Stage 1 in 10 (17.2%), Stage 2A in 7 (12.1%), Stage 2B in 5 (8.6%), Stage 3 in 22 (37.9%), Stage 4 in 4 (6.9%) and Stage 4S in 10 (17.2%) children. Thirty-three (56.9%) received chemotherapy while 5 (8.6%) had a laminectomy as part of surgical procedure. Twenty-seven (46.6%) received radiotherapy (RT). Beam energy was 1.25 MV in 11 (41%), 250 KV in 10 (37%), 4 MV in 4 (15%) and 6 MV photons in 1 patient. One patient received 300cGy in 1 fraction total skin RT using 6 MeV electrons. For the remaining patients, fraction size was 100 cGy in 6 (22%), 150–180 cGy in 11 (41%), 200 cGy in 4 (15%) and 250–300 cGy in 3. Three patients had total body irradiation at 333cGy for 3 fractions. For all children who received RT, median total dose was 2000 cGy (range, 300–3900 cGy). Patients who had RT had plain films of the irradiated area every 1 to 2 years until at least the age of puberty. Median follow-up was 10 years (range 5–46 years). The overall 5-, 10- and 15-year scoliosis-free rates were 87.6%, 79.0% and 76.0% respectively. Twelve (21%) developed scoliosis at a median time of 51 months (range, 8–137 months) from initial diagnosis. The degree of scoliosis was mild (less than or equal to 20 degrees) in 8 (67%). Four patients had scoliosis ranging from 30 to 66 degrees; three of these patients required orthopedic intervention, while 1 had an underlying Duchenne muscular dystrophy which manifested itself 8 years after diagnosis. Median time to scoliosis was 23 months in children who had a laminectomy (range 8–54 months). On multivariate analysis, both history of laminectomy (p = .0005) and use of RT (p = .0284) were found to be risk factors for development of scoliosis. Gender, age at diagnosis, INSS stage, primary site and use of chemotherapy were not found to be significant. Both fraction size and beam energy were also not significant, but increasing total RT dose was found to be significant on logistic regression analysis (p = 0.0017).The 15-year scoliosis-free rates were 20% for children who had a laminectomy and 81.3% for those who did not have laminectomy. The 15-year scoliosis-free rates for children treated with RT doses 0 cGy, 1–1750 cGy, 1751–2300 cGy and greater than 2300 cGy were 91.7%, 87.5%, 51.4% and 44.4% respectively (p = .0039,log-rank). Treatment-related factors, namely laminectomy and radiotherapy, were found to increase the risk of scoliosis in patients with neuroblastoma. Children who had a laminectomy were more likely to manifest scoliosis earlier. Increasing RT dose was found to impact adversely on the development of scoliosis.