Malignant pleural mesothelioma (MPM) is difficult to distinguish from reactive mesothelial proliferations (RMPs). It is uncertain whether miRNAs are useful biomarkers for differentiating MPM from RMPs. Thus, we screened with a quantitative RT-PCR (RT-qPCR)-based platform the expression of 742 miRNAs in formalin-fixed, paraffin-embedded, preoperative diagnostic biopsy samples, surgically resected MPM specimens previously treated with chemotherapy, and corresponding non-neoplastic pleura (NNP), from five patients. miR-126, miR-143, miR-145, and miR-652 were significantly down-regulated (≥twofold) in resected MPM and/or chemotherapy-naïve diagnostic tumor biopsy samples. The miRNA expression pattern was validated by RT-qPCR in a cohort of 40 independent MPMs. By performing binary logistic regression on the RT-qPCR data for the four miRNAs, the established four-miRNA classifier differentiated MPM from NNP with high sensitivity and specificity (area under the curve, 0.96; 95% CI, 0.92-1.00). The classifier's optimal logit(P) value of 0.62 separated NNP and MPM samples with a sensitivity of 0.95 (95% CI, 0.89-1.00), a specificity of 0.93 (95% CI, 0.87-0.99), and an overall accuracy of 0.94 (95% CI, 0.88-1.00). The level of miR-126 in MPM was inversely correlated with that of the known target, the large neutral amino acid transporter, small subunit 1 (r = -0.38; 95% CI, -0.63 to -0.06). Overall, these results indicate that these four miRNAs may be suitable biomarkers for distinguishing MPM from RMPs.