Introduction. The high mortality rate in patients with lung cancer (LC) is due to the lack of highly sensitive diagnostic markers of this disease. Genetic and epigenetic alterations in tumor cells, for example, aberrant microRNA expression, can be proposed. It is known that extracellular/circulating microRNA of biological fluids, in complexes with proteins, or packaged in extracellular vesicles is of interest for the diagnosis of tumor diseases.Aim. To perform a comparative analysis of miRNA expression in plasma and plasma extracellular vesicles of LC patients and healthy donors. Based on the obtained results, to propose a diagnostic panel to identify patients with LC.Materials and methods. Blood plasma was obtained from blood samples of healthy donors and LC patients by sequential centrifugation. Then, a fraction of extracellular vesicles (40–150 nm in size) was isolated from a part of the obtained plasma supernatant by the method of aggregation-precipitation with polyethylene glycol/blue dextran. MicroRNAs were isolated from both blood plasma fractions of patients and healthy donors using guanidine isothiocyanate and octanoic acid. Expression of 17 miRNAs most characteristic for the development of LC according to our and literature data in the above-mentioned blood plasma fractions was analyzed by stem-loop reverse transcription polymerase chain reaction.Results. 29 and 10 miRNA pairs were differentially expressed in plasma extracellular vesicles and plasma of lung cancer patients and donors. Thus, plasma extracellular vesicles are characterized by greater potential as a source for miRNA based lung cancer diagnostic panels in comparison with blood plasma. Diagnostic algorithm based on aberrant miRNA expression of 8 different miRNAs (miRNA-30e, -1, -125b, -133, -222, -374, -425, -660) composed in 6 pairs was designed. This algorithm allows to diagnose 100 % of patients with lung cancer stages II–IV.Conclusion. Extracellular plasma vesicles represent a promising source of diagnostically significant microRNAs compared to plasma microRNAs. For the diagnosis of patients with non-small cell lung cancer with 100 % sensitivity and specificity, a panel of 8 microRNAs (6 miRNA pairs) was proposed.