Surgical management of severe rigid dystrophic neurofibromatosis (NF) scoliosis is technically demanding and produces varying results. In the current study, we reviewed 9 patients who were treated with combined anterior and posterior fusion using different types of instrumentation (i.e., pedicle screw, hybrid, and all-hook constructs) at our institute.Between September 2001 and July 2010 at our institute, 9 patients received anterior release/fusion and posterior fusion with different types of instrumentation, including a pedicle screw construct (n = 5), a hybrid construct (n = 3), and an all-hook construct (n = 1). We compared the pedicle screw group with the hybrid group to analyze differences in preoperative curve angle, immediate postoperative curve reduction, and latest follow-up curve angle.The mean follow-up period was 9.5 ± 2.9 years. The average age at surgery was 10.3 ± 3.9 years. The average preoperative scoliosis curve was 61.3 ± 13.8°, and the average preoperative kyphosis curve was 39.8 ± 19.7°. The average postoperative scoliosis and kyphosis curves were 29.7 ± 10.7° and 21.0 ± 13.5°, respectively. The most recent follow-up scoliosis and kyphosis curves were 43.4 ± 17.3° and 29.4 ± 18.9°, respectively. There was no significant difference in the correction angle (either coronal or sagittal), and there was no significant difference in the loss of sagittal correction between the pedicle screw construct group and the hybrid construct group. However, the patients who received pedicle screw constructs had significantly less loss of coronal correction (P < .05). Two patients with posterior instrumentation, one with an all-hook construct and the other with a hybrid construct, required surgical revision because of progression of deformity.It is difficult to intraoperatively correct dystrophic deformity and to maintain this correction after surgery. Combined anterior release/fusion and posterior fusion using either a pedicle screw construct or a hybrid construct provide similar curve corrections both sagittally and coronally. After long-term follow-up, sagittal correction was maintained with both constructs. However, patients treated with posterior instrumentation using pedicle screw constructs had significantly less loss of coronal correction.