Background: Mounting studies report that microRNAs may be useful as diagnostic biomarkers of gastric cancer (GC), and compared with healthy individuals, gastric cancer patients tend to have a significantly high concentration of microRNA-106a (miR-106a) in plasma. Accordingly, this meta-analysis aims to evaluate the potential diagnostic role of miR-106a in gastric cancer. Methods: PubMed, Embase, Web of Science and the Cochrane Library were searched for publications, which are related to the diagnostic role of miR-106a in gastric cancer. Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) was used to estimate the quality of all the studies. Then, the data of each qualified study were extracted to perform meta-analysis. The overall test performance was checked by the Receiver operating characteristic curves (ROC). And the Chi-square and I-2 test were used to assess the existence of heterogeneity. Results: Four studies involving 238 GC patients and 139 controls were included in our meta-analysis. The results indicated that the pooled sensitivity and specificity are 71.6% (95% confidence interval (CI): 59.6%-81.2%) and 79.2% (95% CI: 55.3%-92.1%) respectively. Furthermore, the value of AUC (area under the summary ROC curve) is 0.79. Conclusion: This meta-analysis indicates that miR-106a has a moderate sensitivity and specificity in the diagnosis of GC. In order to improve the accuracy of diagnosis, more large-scale studies about the role of miR-106a are required.