BACKGROUND: Circulating microRNAs (miRNAs) are now used as non-invasive diagnostic indicators in a variety of malignancies. Our objective is to use a meta-analysis to assess the diagnostic performance of circulating miRNAs in gastric cancer. METHODS: We reviewed databases and methodically obtained papers for analysis up to October 15th, 2021. The random-effect meta-analysis was performed to construct pooled diagnostic parameter. To detect the causes of heterogeneity, spearman threshold effect analysis and subgroup analysis were performed. The I2 and Chi-square test were also used to examine the heterogeneity. The subgroup analyses were conducted based on sample types (serum/plasma/blood), normalized genes (U6, miR-16, and miR-39), qPCR mastermix (SYBR and Taqman), and country. Finally, the publication bias was estimated using Egger's funnel plot asymmetry test. RESULTS: A total of 40 articles covering 73 studies (59 microRNAs) were included, containing 11,022 participants (6,324 cases and 4,698 controls). The overall pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and area under the curve (AUC) were 0.75 (95% CI: 0.74-0.77), 0.79 (95% CI: 0.78-0.80), 4.081 (95% CI: 3.43-4.85), 0.28 (95% CI: 0.25-0.32), 16.08 (95% CI: 12.34-20.95), and 0.877 (CI: 0.84- 0.90), respectively. We conducted a subgroup analysis of diagnostic value, which revealed that serum type, U6 reference gene, SYBR mastermix, and East Asian Countries (China, Japan, Taiwan) had better diagnostic value. CONCLUSIONS: In consequence, circulating miRNAs can serve as diagnostic biomarkers for gastric cancer. However, specific miRNAs still need to be discovered in diagnosing gastric cancer, especially early screening.