Abstract: To provide reference for the prevention, diagnosis and treatment of diabetes by means of comprehensively analyzing the relationship between serum ferritin level and diabetes. The software of review manager 5 (Revman 5) was applied to analyze the data for meta analysis. Based on the screening of the criteria about literature inclusion and exclusion, there were 23 case-control studies eventually included in the Meta-analysis, which referred to 3,269 subjects, including 1,888 diabetic cases and 1,808 non-diabetic cases. In terms of the meta-analysis, the D-L method (DerSimonian & Laird) of random effects model was used in the statistical analysis data because the heterogeneity test results suggested that the heterogeneity between studies was significant. The effect size showed that... the serum ferritin levels in patients with diabetes was higher than that in non-diabetic patients and the difference represented the statistical significance (WMD = 105.60, 95% CI (87.23, 123.97), P<0.00001). The results remained unchanged after the sensitivity analysis of the study. It also showed a good symmetry of the funnel diagram in this study with the fail-safe factor of Nfs0.05=3874, suggesting that this study was less affected by the publication bias and indicating the stable and reliable results of meta-analysis. Serum ferritin level was closely related to diabetes and may be one of the important risk factors of diabetes.