OBJECTIVE: The aim of this study was to comprehensively assess the overall diagnostic value of circulating microRNAs (miRNAs or miRs) as biomarkers for the early diagnosis of diabetic kidney disease (DKD) through Meta-analysis, and to identify potential molecular biomarkers with higher diagnostic value for early DKD. METHODS: The CNKI, Wanfang date, VIP, Pubmed, Embase, Web of Science, and Cochrane Library until January 2024 were searched. Relevant studies associated with the value of miRNAs in the diagnosis of early DKD were selected. Case numbers, sensitivity, and specificity were extracted from the included literature for both the observation and control groups. RESULTS: Nine studies including 655 cases of early DKD patients and 664 cases as a control group were conducted. The comprehensive sensitivity was 0.76, comprehensive specificity was 0.74, combined positive likelihood ratio was 2.9 and the combined negative likelihood ratio was 0.33, diagnostic odds ratio (DOR) was 9. The summary receiver operating characteristic (SROC) curve was drawn and the area under the curve (AUC) was 0.79. Blood and urine source data were analyzed and showed that urine source miRNA had a higher sensitivity (0.82vs 0.68) and a higher DOR (10.5vs 8.2) than blood source miRNA. CONCLUSION: MiRNAs may serve as promising noninvasive biomarkers for the early diagnosis of DKD. The diagnostic value of miRNAs in urine samples may be higher than that in blood samples. The combined detection of some miRNAs or other clinical indicators can enhance the accuracy of early DKD diagnosis. SYSTEMATIC REVIEW REGISTRATION: https://osf.io, identifier DOI: 10.17605/OSF.IO/FC6DK.