Background & Aims: Growing evidence has shown that M2-PK is involved in cancer diagnosis and prognosis. The overall diagnostic accuracy of the pyruvate kinase isoenzyme type M2 (M2-PK) in biliary tract carcinoma (BTC) remains controversial. We performed a meta-analysis to evaluate the diagnostic value of M2-PK for BTC.Methods: The online PubMed, Cochrane, Web of Science, and Embase databases were searched for eligible studies published until August 8th, 2017. The Quality Assessment for Diagnostic Accuracy Studies 2 (QUADAS-2) was used to evaluate study quality. All statistical analyses were conducted with Stata 12.0.Results: We included 7 studies from 5 articles with 410 patients with BTC and 438 controls. The pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and AUC for M2-PK in the diagnosis of BTC were 0.79 (95%CI 0.70-0.86), 0.81 (95%CI 0.71-0.88), 4.1 (95%CI 2.5-6.8), 0.26 (95%CI 0.16-0.41), 17.159 (95%CI 5.468-54.071), and 0.87 (95%CI 0.83-0.89), respectively. The same indicators assessed for CA19-9 were as follows: 0.70 (95%CI 0.62-0.77), 0.71 (95%CI 0.45-0.87), 2.38 (95%CI 1.2-4.73), 0.43 (95%CI 0.34-0.53), 6.28 (95%CI 2.4-16.44) and 0.73 (95%CI 0.69-0.77), respectively. Additionally, the diagnostic value of M2-PK varied based on characteristics of golden methods and different cut-off values.Conclusions: This meta-analysis showed that M2-PK had a better diagnostic accuracy for BTC compared with CA19-9, with moderate diagnostic performance. However, prospective studies are required to confirm its diagnostic value.