Objective To systematically evaluate the clinical efficacy and late effect of group cognitive behavioral therapy（ GCBT） in the treatment of insomnia. Methods EMbase， Cochrane Library， Medline， CNKI and Wanfang database were searched for randomized controlled trials that met the inclusion criteria. The GCBT group received group cognitive behavior therapy， including sleep health education， cognitive therapy， relaxation training， sleep restriction and stimulation control. The control group received placebo treatment， health education， routine care， waiting for treatment. The questionnaire of sleep latency， wake-up time after sleep， total sleep time and sleep efficiency， sleep severity index and Pittsburgh sleep quality index were extracted from sleep diary， and meta-analysis was conducted by RevMan 5.3 and STATA 15.1 software. Results Finally， 11 articles were included， a total of 814 patients， including 441 cases in GCBT group and 373 cases in control group. The results of meta-analysis showed that： sleep latency（ WMD=-15.06，95%CI： -19.06- -11.05，P＜0.05）， awakening time after sleep（ WMD=-34.95，95%CI：-49.96- -19.93，P＜0.05）， insomnia severity index（ WMD=-6.13，95%CI：-8.04- -4.22，P ＜ 0.05）， Pittsburgh sleep quality index （WMD=-2.49，95%CI：-4.11- -0.87，P ＜ 0.05） of the GCBT group were all lower than those of the control group， while the sleep efficiency of the GCBT group was higher than that of the control group（ WMD=10.46， 95%CI：6.89-14.03，P ＜ 0.05）. There was no significant difference in the total sleep time between the two groups（ P ＞ 0.05）. At early follow-up， the awakening time（ WMD=-32.51， 95%CI：-58.61- -6.41，P ＜ 0.01）， insomnia severity index（ WMD=-6.01，95%CI：-8.40- -3.62，P＜0.05）， Pittsburgh sleep quality index （WMD=-4.33，95%CI：-7.06- -1.59，P ＜ 0.05） were lower in GCBT group than those in control group， and total sleep time（ WMD=0.32， 95%CI：0.10-0.54，P ＜ 0.05）， sleep efficiency（ WMD=10.51， 95%CI：5.99- 15.02，P ＜ 0.05） were higher than those in the control group， and there was no significant difference in the sleep latency between the two groups（ all P＞0.05）. At long-term follow-up， the sleep efficiency of GCBT group was higher than that of control group（ WMD=5.30，95%CI：1.61-8.98，P＜0.05）， and Pittsburgh sleep quality index score was lower than that of control group（ WMD=-2.80，95%CI：-3.82- -1.78，P＜0.05）. There was no significant difference between the two groups in sleep latency， wake-up time and total sleep time（ all P＞0.05）. Conclusions The clinical efficacy of GCBT in patients with insomnia is reliable， and it will continue to be effective in the late follow-up， but its efficacy will decline with time prolongation.