OBJECTIVE： To evaluate the efficacy and safety of semaglutide in the treatment of type 2 diabetes mellitus， and to provide evidence-based reference in clinic. METHODS： Retrieved from Wangfang， CNKI， PubMed， Embase， Medline and the Cochrane library， randomized controlled trials（RCTs） about semaglutide alone or combined with other hypoglycemic drugs （trial group） versus placebo or other hypoglycemic medicine alone or combination （control group） in the treatment of type 2 diabetes  mellitus were collected. After data extraction and literature quality evaluation， Meta-analysis was performed by using Rev Man 5.3 software. RESULTS：Totally 7 literatures were included， involving 7 708 patients. Results of Meta-analysis showed that compared with control group， the level of glycosylated hemoglobin [vs. placebo： MD＝－1.48， 95％CI（－1.68，－1.28）， P＜0.001； vs. positive control medicine：MD＝－0.95， 95％CI（－1.17，－0.73）， P＜0.001]， fasting plasma glucose [vs. placebo： MD＝－1.87， 95％CI（－2.25，－1.50）， P＜0.001； vs. positive control medicine： MD＝－1.07， 95％CI（－1.58，－0.55）， P＜0.001]， body weight [vs. placebo： MD＝－3.15， 95％CI（－3.98，－2.32）， P＜0.001； vs. positive control  medicine： MD＝－3.64， 95％CI（－4.60，－2.69）， P＜0.001] and body weight index [vs. placebo： MD＝－1.11， 95％CI（－1.40，－0.81）， P＜0.001； vs. positive control medicine： MD＝－1.34，95％CI（－1.67，－1.00）， P＜0.001] in trial group were decreased significantly. For safety， the incidence of nausea and diarrhea of somaglutide was significantly higher than control group（P＜0.05）. CONCLUSIONS：Semaglutide showes superiority in reducing the levels of glycosylated hemoglobin， fasting plasma glucose， body weight and body weight index in patients with type 2 diabetes mellitus， but the occurrence of nausea and diarrhea in patient should be monitored.