Objective：To evaluate the safety and effectiveness of D2 lymphadenectomy and D3 lymphadenectomy in the treatment of gastric cancer. Methods：PubMed，Cochrane Library，Chinese biomedicine literature database，Chinese journal full-text database and Chinese scientific journal full-text database were searched to identify randomized controlled trials（RCTs） comparing D2 with D3 lym－phadenectomy for gastric cancer. We also hand-searched related key Chinese journals to identify potential studies. Quality of the included studies was assessed and Meta analysis was performed using RevMan 5.1 software. Results：Six RCTs were included. Meta analysis showed that there was no significant difference between D2 group and D3 group in terms of postoperative mortality（odds ratio（OR）=1.01，95％confidence interval（95％CI）=0.44 to 2.32），incidence of overall postoperative complications（OR=1.53，95%CI=0.81 to 2.89），five-year overall survival rate（OR=1.08，95%CI=0.81 to 1.44） and recurrence rate（OR=0.89，95%CI=0.67 to 1.81）（P=0.980，0.190，0.590，0.410）. D3 group had higher incidence of postoperative lymphatic fistula（OR=0.66，95%CI=0.01 to 0.50）and in－tractable diarrhea（OR=0.11，95%CI=0.01 to 0.90），but D2 group had higher pulmonary infection rate（OR=3.71，95%CI=1.43 to 9.61） when doing comparison with statistical differences（P=0.009，0.040，0.007）. No difference was found between two groups regarding incidence of anastomotic leakage，pancreatic fistula，abdominal abscess，upper gastrointestinal bleeding，wound infection，reoperation rate and intestinal obstruction（P=0.240，0.500，0.080，0.400，0.710，0.570，0.960）. D3 group had longer operation time（weighted mean difference（WMD）=-154.86，95%CI=1.04 to 4.18） and more intraoperative blood loss voulme（WMD=301 ml，95％CI=151.55 to 450.45） compared with those of D2 group（P=0.008，0.000）. No statistically significant difference was existed in mean intraoperative blood transfusion volume（WMD=-138.47 ml，95%CI=-379.96 to -103.01） and hospitalization time（WMD=1.77 d，95%CI=-1.65 to 5.15）between groups（P=0.31，0.26）. Conclusions：Current evidence suggests that D3 lymphadenectomy has no survival advantage and has higher operation risk when compared with D2 lym－phadenectomy，therefore is not recommend for routine use.