Objective: To evaluate the effect of functional training on the improvement of Cobb Angle, ATR and quality of life in adolescent idiopathic scoliosis. Methods: Database of CNKI, WanFang, WeiPu, Web of Science, PubMed were reconnoitered for the purpose of research articles of Interest. Literature was screened according to the inclusion criteria, randomized controlled trials (RCTS) were collected and Cochrane risk bias assessment methodology quality was used, and Revman 5.3 software was used to conduct combined effector subgroup analysis on outcome indicators (Cobb Angle, ATR, SRS-22 patient questionnaire). Results: A total of 10 RCT (398 AIS patients) were included in the study. There were two high , five moderate and three low quality studies, respectively. Compared with other non operative treatments, functional training could reduce Cobb horn in patients, MD = -6.56 (95% CI =-7.30--5.83, P <0.01, I 2=0). Further subgroup analysis showed that there were no statistically significant differences in effect size between the subgroups of Cobb Angle, age, intervention period and control method at the time of inclusion ( P >0.05). Compared with the non intervention blank group, the functional training group could reduce the Cobb Angle of patients ( MD = -5.25 , 95% CI =-8.27- -2.98 , P < 0.01 ), with high heterogeneity ( I 2=61%). Effect on angle of trunk rotation(ATR):functional training group could reduce patients ATR, MD =-1.91 (95% CI =-2.25--1.57, P <0.01, I 2=0). The impact on the quality of life:functional training of patients with SRS-22 questionnaire function, pain had no statistical significance ( P >0.05), and self image ( MD =0.64, 95% CI =0.53- 0.75 , P <0.01), mental health ( MD =0.44, 95% CI =0.15-0.74, P <0.01), satisfaction ( MD = 0.58 , 95% CI =0.11- 1.06 , P = 0.02 ) were statistically significant. Conclusion Functional training can not only reduce the Cobb Angle and ATR of AIS patients, but also improve the quality of life in terms of self image, mental health and satisfaction of the patients. However, due to heterogeneity and bias, more high quality studies are needed to verify.