Objective: To systematically evaluate the effects of Tai chi for improving elderly patients with type 2 diabetes. Methods: According to PRISMA checklist, we conducted this standard meta-analysis. The multiple databases like Pubmed, Embase, and Cochrane databases were used to search for the relevant studies, and full-text articles involved in the evaluation of Tai chi in improving elderly patients with type 2 diabetes. Review manager 5.2 was adopted to estimate the effects of the results among selected articles. Forest plots, sensitivity analysis and funnel plot for the articles included were also conducted. Results: Finally, 7 relevant studies were eventually satisfied the included criteria. We found that Tai chi group had lower glucose than control group (mean difference (MD)=-12.47, 95%CI [-21.20, -3.73], P=.005; I2 = 32%), Tai chi group had higher activities-specific balance confidence (ABC) scale than control group (MD =9.26 with 95%CI [6.68, 11.83], P < .001) and Tai chi group had higher single limb standing test score than control group (MD = 8.38, 95%CI [4.02, 12.74], P = .001). The study was robust and limited publication bias was observed in this study. Conclusion: Since we found Tai chi had better performance than usual care in improving old diabetes patients' glucose and life quality, the study supports that Tai chi can help old diabetes patients from several aspects including disease indicators, independence and life quality.