Objective To investigate the safety and accuracy of robot-assisted pedicle screw implantation in the adolescent idiopathic scoliosis (AIS) surgery. Methods The clinical data of 46 patients with AIS who were treated with orthopedics, bone graft fusion, and internal fixation via posterior approach between June 2018 and December 2019 were analyzed retrospectively. Among them, 22 cases were treated with robot-assisted pedicle screw implantation (robot group) and 24 cases with manual pedicle screw implantation without robot assistance (control group). There was no significant difference in gender, age, body mass index, Lenke classification, and preoperative Cobb angle of the main curve, pain visual analogue scale (VAS) score, Japanese Orthopaedic Association (JOA) score between the two groups (P>0.05). The intraoperative blood loss, pedicle screw implantation time, intraoperative pedicle screw adjustment times, and VAS and JOA scores after operation were recorded. The Cobb angle of the main curve was measured on X-ray film and the spinal correction rate was calculated. The screw position and the accuracy of screw implantation were evaluated on CT images. Results The operation completed successfully in the two groups. The intraoperative blood loss, pedicle screw implantation time, and pedicle screw adjustment times in the robot group were significantly less than those in the control group (P<0.05). There was 1 case of poor wound healing in the robot group and 2 cases of mild nerve root injury and 2 cases of poor incision healing in the control group, and there was no significant difference in the incidence of complications between the two groups (P=0.667). All patients in the two groups were followed up 3-9 months (mean, 6.4 months). The VAS and JOA scores at last follow-up in the two groups were superior to those before operation (P<0.05), but there was no significant difference in the difference of pre- and post-operative scores between the two groups (P>0.05). The imaging review showed that 343 screws were implanted in the robot group and 374 screws in the control group. There were significant differences in pedicle screw implantation classification and accuracy between the two groups (89.5% vs 79.1%)(Z=−3.964, P=0.000; χ2=14.361, P=0.000). At last follow-up, the Cobb angles of the main curve in the two groups were significantly lower than those before operation (P<0.05), and there was significant difference in the difference of pre- and post-operative Cobb angles between the two groups (t=0.999, P=0.323). The spinal correction rateswere 79.82%±5.33% in the robot group and 79.62%±5.58% in the control group, showing no significant difference (t=0.120, P=0.905). Conclusion Compared with manual pedicle screw implantation, robot-assisted pedicle screw implantation in AIS surgery is safer, less invasive, and more accurate.