OBJECTIVE：To systematically evaluate the occurren ce of non-immune related adverse events （AEs）caused by immune checkpoint inhibitors （ICIs）alone or combined with routine chemotherapy in the treatment of non-small cell lung cancer （NSCLC），and to provide evidence-based reference for clinical medication. METHODS ：Retrieved from PubMed ，Cochrane Library，Embase，CNKI，CBM，VIP and Wanfang database during the inception to Oct. 2020，randomized controlled trials （RCT） about ICIs alone or combined with routine chemotherapy （trial group ）versus routine chemotherapy or placebo combined with routine chemotherapy （control group ） were collected. After literature screening and data extraction ，the quality of included literatures were evaluated with bias risk evaluation tool recommended by Cochrane systematic evaluator manual 5.1.0. Meta-analysis was performed by using Rev Man 5.3 software and Stata 15.0 software. Sensitivity analysis was conducted with Stata 15.0 software. Inverted funnel plot and Egger ’s test were used to analyze publication bias. RESULTS ：A total of 20 RCTs were included ， involving 12 283 patients. Results of Meta-analysis showed that the incidence of all grades and s evere AEs ，anemia，neutropenia， vomiting and alopecia as well as the incidence of thrombocytopenia，nausea and peripheral neuropathy in all grades of trial group were all significantly lower than control com group（P＜0.05）. There was no statistical significance in the incidence of termination of treatment ， death， severe thrombocytopenia， severe nausea and severe peripheral neuropathy or all grades and severe diarrhea between 2 groups（P＞0.05）. Subgroup analysis showed that the incidence of all grade and total severe AEs ，the incidence of anemia ，neutropenia，thrombocytopenia，clinically relevant symptoms （except for severe diarrhea），termination of treatment and death of patients receiving ICIs alone in trial group were significantly lower than control group（P＜0.05）. The incidence of ermination of treatment and death ，the incidence of nausea ，vomiting，diarrhea and alopecia in all grade ，severe diarrhea of patients receiving ICIs and chemotherapy in trial group were all significantly higher than control group （P＜0.05）. Sensitivity analysis supported the above results. Analyze publication bias results showed that the possibility of publication bias in this study was small. CONCLUSIONS ：For NSCLC patients ，the safety of ICIs is better than that of routine chemotherapy or placebo combined with routine chemotherapy in the treatment-related AEs ，hematologic toxicity and clinically relevant symptoms ；however，the risks of treatment discontinuation ，AEs-induced deaths ，and all-grade nausea ，vomiting， diarrhea，alopecia and severe diarrhea will be increased in the ICIs combined with routine chemotherapy.