Thyroid cancer (TC) ranks as the most prevalent malignant endocrine neoplasm worldwide, holding 9th place in terms of occurrence and ranking 24th for fatality among all diagnosed malignant tumors. Thus, discovering efficient biomarkers for timely identification and diagnosis of TC is pivotal. A meta-analysis was conducted to evaluate the diagnostic capability of circulating microRNAs (miRNAs) in detecting TC, registered under INPLASY202360048 on the INPLASY website. A systematic search of four databases (PubMed, Embase, Web of Science, and Cochrane Library) was performed to identify relevant articles published from inception until December 22, 2022. Stata 14.0 software was used to calculate the pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic ratio (DOR), and area under the summary receiver operating characteristic (ROC) curve to assess the accuracy of miRNAs in the diagnosis of TC. Study heterogeneity was quantitatively evaluated using the Cochran-Q test and I2 statistic. Given the significant variability among studies, we opted for a random-effect model. Subgroup analysis and regression analysis were conducted in an effort to identify any possible factors contributing to heterogeneity. Our meta-analysis included 935 TC patients and 914 non-TC controls across 48 studies from 15 articles. The results showed that the summary sensitivity and specificity were 0.80 (95% confidence interval [CI]: 0.76, 0.84) and 0.81 (95% CI: 0.77, 0.85), respectively, the combined positive likelihood ratio was 4.27 (95%CI:3.43,5.33), the negative likelihood ratio was 0.24 (95%CI:0.20,0.30), and the diagnostic ratio was 17.55(95%CI:12.26,25.12). The combined AUC is 0.88 (95%CI: 0.84, 0.90). MiRNA profiling, regulation mode of miRNAs, and cut-off value settings emerged as primary heterogeneity sources. Based on our meta-analysis, circulating miRNAs show promise as a non-invasive diagnostic biomarker for TC.