OBJECTIVE： To systematically evaluate therapeutic efficacy of Xiaoke formula in the treatment of type 2 diabetes mellitus （T2DM） with Qi and Yin deficiency， and to provide evidence-based reference in clinic. METHODS： Retrieved from PubMed， Embase， Cochrane Library， CBM and CNKI， randomized controlled trials （RCTs） about modified Xiaoke formula （trial group） with Qi and Yin deficiency versus Metformin tablet or insulin （control group） in the treatment of T2DM were included. After literature screening and data extraction， the qualities of included studies were evaluated by modified Jadad scoring scale. Meta-analysis was performed by using Rev Man 5.3 statistical software. RESULTS： Totally 18 RCTs were included， involving      2 065 patients. Meta-analysis showed that response rate of trial group [OR＝2.89， 95％CI（2.16，3.86）， P＜0.001] was significantly higher than that of control group， while fasting blood glucose [SMD＝－0.72， 95％CI（－0.96，－0.47）， P＜0.001]， 2 h postprandial blood glucose [SMD＝ －0.72， 95％CI（－0.96，－0.47）， P＜0.001]， glycosylated hemoglobin [MD＝－0.65， 95％CI（－0.88，－0.42）， P＜0.001] and TCM syndrome score [MD＝－5.77， 95％CI（－7.46，－4.08）， P＜0.001] were significantly lower than control group. Totally 18 RCTs were included， involving 18 formula and 66 ingredients. The sum of use frequency of various medicinal materials was 218. Trichosanthes， Coptis chinensis， Ophiopogon japonicus and so on were used most frequently. CONCLUSIONS： Xiaoke formula shows significant therapeutic efficacy for T2DM with Qi and Yin deficiency， and reduces related index levels and TCM syndrome score effectively.