Fine-needle aspiration (FNA) is routinely used in the preoperative evaluation of thyroid nodules. However, approximately 5–20% of thyroid nodules are considered indeterminate or suspicious cases that do not meet clinical standards. The B-RAFV600E mutation has been reported in FNA specimens. We conducted a systematic review to evaluate the diagnostic value of testing for B-RAFV600E in thyroid nodules that are difficult to diagnose by FNA. A systematic literature search was performed from January 1, 2002 to June 30, 2012. Articles were obtained by searching two electronic databases (MEDLINE and EMBASE), hand searching selected journals, and contacting authors. Article quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool. Sensitivity, specificity, and other measures of accuracy were pooled using random effects models. Summary receiver operating characteristic (SROC) curves were used to summarize overall diagnostic accuracy. A total of 16 studies incorporating 1131 patients were included in a meta-analysis on diagnostic accuracy of B-RAFV600E tests. Pooled sensitivity was 0.60 (95% confidence interval [CI]: 0.556–0.634), pooled specificity was 0.99 (95% CI 0.976–0.997), and the area under the curve of the SROC curve was 0.8376. Q index value was 0.7696. Our data suggest a potentially useful adjunct to evaluating thyroid nodules that are difficult to diagnose. The B-RAFV600E test has a high positive predictive value and could help clinicians formulate a more individualized treatment schedule. When supplemented with other noninvasive test methods, the B-RAFV600E test could be a powerful adjunct with extensive clinical applications. Diagn. Cytopathol. 2014;42:94–101. © 2013 Wiley Periodicals, Inc.