Objective: To systematically evaluate the effect of metformin treatment for diabetes on patients' cognitive condition. Methods: The databases of CBM、CNKI and PubMed were searched to collect randomized controlled trials of metformin for the treatment of diabetes mellitus published between the date of establishment and December 31, 2022, and the data were statistically analyzed by RevMan 5.3 software, according to MMSE scores, MoCA scores, NIHSS scores and adverse effects were analyzed. Results: These 10 papers contain 1328 patients, and there are 663 patients in the treatment group while 665 patients in the control group. Meta-analysis showed that the MMSE scores in the observation group of sulfonylureas were higher than those in the control group, and there was significant difference, but in the insulin group the MMSE scores in the observation group were lower than those in the control group, and the difference was statistically significant. The difference between the MoCA scores and the control group was not statistically significant; the NIHSS scores of treatment group were significantly higher than those in the control group, and the difference between groups was statistically significant; and the adverse effects between the treatment group and the control group were statistically significant. No statistically significant difference was observed in adverse events between two groups. Conclusion: Metformin can improve the cognitive function and promote the recovery of neurological function in diabetic patients. However, no significant improvement in cognitive function was observed compared with insulin. Metformin has a good safety profile, so it can be used in the treatment of diabetes mellitus with cognitive impairment.