To evaluate the role and value of Changhai fulcrum bending radiograph(CH-FBR) in curve flexibility assessment of adolescent idiopathic scoliosis(AIS) patients. Thirty-seven AIS patients treated between June 2012 and August 2013 were enrolled, including 31 female and 6 male patients whose age ranged from 10 to 19 years, averaged of 15.0 years. The assessment of radiographs included preoperative standing posterior-anterior radiograph, supine side-bending radiograph, traditional fulcrum bending radiograph, Changhai fulcrum bending radiograph and postoperative standing posterior-anterior radiograph. Postoperatively, radiographs were assessed at one week. The CH-FBR was performed at the lowest height and the optimized height which means the weight on the fulcrum touch the maximum. All measurements of angle were made with use of the Cobb method. The flexibility of the curve as well as the correction rate and fulcrum bending correction index (FBCI) were calculated for all patients. The maximum height of CH-FBR, basic weight and maximum weight were measured for all AIS. Paired t-tests were used to assess differences between preoperative and postoperative curves within group samples. The Pearson correlation coefficients were calculated using bivariate analysis between CH-FBR flexibility rate and correction rate, the maximum height of CH-FBR and maximum weight, the height changes of CH-FBR and weight changes. A total of 46 curves were involved in this study, including 28 thoracic and 18 thoracolumbar/lumbar curves. Preoperatively, the mean Cobb angle of the 46 structural curves was 47° ± 11°. Postoperatively, the mean Cobb angle was 11° ± 5°. Cobb's angle in supine side-bending(t = 7.2, P = 0.001), traditional fulcrum bending (t = 7.1, P = 0.001) and lowest height of Changhai fulcrum bending (t = 6.5, P = 0.001) were significantly different from the postoperative Cobb angle; Cobb's angle in traditional FBR (t = 11.0, P = 0.001) and lowest height of Changhai fulcrum bending (t = 13.6, P = 0.001) were significantly different from the optimized height CH-FBR Cobb angle. There was no significant difference found between traditional FBR Cobb angle and lowest height CH-FBR Cobb angle (t = 2.0, P = 0.051), optimized height CH-FBR Cobb angle and postoperative Cobb angle (t = 0.9, P = 0.36), lowest height CH-FBR Cobb angle and traditional FBR Cobb angle(t = 2.0, P = 0.051). The maximum height of CH-FBR, basic weight and maximum weight were (29.6 ± 1.4)cm,(20 ± 6)kg, and (40 ± 6) kg. Preoperatively, the mean Cobb angle of the 28 structural curves(main thoracic curves) was 46° ± 11°. Postoperatively, the mean Cobb angle was 12° ± 6°. Preoperatively, the mean Cobb angle of the 18 structural curves(thoracolumbar/lumbar curves) was 49° ± 12°. Postoperatively, the mean Cobb angle was 10° ± 5°. The results were same in 28 structural curves, 18 structural curves as well as 46 curves. Correlation analysis of 46 curves indicated that the maximum height of CH-FBR positively correlated with maximum weight (r = 0.69, r(2) = 0.47, P = 0.001), the height changes of CH-FBR positively correlated with weight changes on CH-FBR (r = 0.62, r(2) = 0.38, P = 0.001). CH-FBR is a more reliable and effective method than traditional FBR and supine side-bending for curve flexibility evaluation in AIS patients. Moreover, compared to the traditional FBR and side-bending radiograph, the flexibility suggested by the optimized height CH-FBR more closely approximates the postoperative result made by pedicle screws fixation and fusion.