Recent studies have shown abnormal microRNA (miRNA) expression levels in the central nervous system (CNS) of cancer patients, suggesting that miRNAs may serve as promising biomarkers for cancers of CNS. However, other studies have arrived at conflicting results. Therefore, this meta-analysis aims to systematically measure the potential diagnostic value of miRNAs for CNS cancers. Electronic databases as well as other sources were searched until to April 12, 2014 for relevant articles. Data from different studies were pooled using the random-effects model. The pooled sensitivity, specificity, positive likelihood ratio (PLR), negative LR (NLR), diagnostic odds ratio (DOR), together with the summary receiver operator characteristic (SROC) curve, and area under the SROC curve (AUC) value were used to estimate overall diagnostic performance. Twenty-three studies from 6 articles were included in the current meta-analysis with a total of 299 CNS cancer patients and 418 controls. The pooled sensitivity, specificity, PLR, NLR, DOR, and AUC were 0.85 (95% CI, 0.80-0.89), 0.83 (95% CI, 0.76-0.88), 5.1 (95% CI, 3.4-7.5), 0.18 (95% CI, 0.12-0.26), 28 (95% CI, 14-58), and 0.91 (95% CI, 0.88-0.93), respectively. Subgroup analyses showed that cerebrospinal fluid (CSF)-based miRNAs assays yielded more accurate results and seemed to be more sensitive in diagnosing of primary central nervous system lymphoma (PCNSL). In conclusion, miRNAs may be suitable for serving as noninvasive biomarkers for CNS cancers detection. However, further validation based on a larger sample of patients and controls is still required.