Sarcoidosis, known as Besnier, Boeck and Schaumann disease (BSS) is a disorder of unknown etiology, in which non-caseating granulomas develop in many organs. The lungs and lymph nodes are affected in 90% of patients. Diagnosis is based on the clinical symptoms, chest x-ray and computerized tomography. It is necessary to find non-caseating epithelioid cell granulomas by biopsy of an involved organ. The diagnosis of sarcoidosis needs differentiation with diseases presenting with similar clinical signs, radiological characteristics and cyto-histopathological evidence, including tuberculosis. The course of the diseaseis usually benign, most often there are spontaneous remissions, however a chronic course as well as recurrence may occur. In extreme cases sarcoidosis may lead to lung fibrosis, which may require lung transplantation. Treatment is indicated in approximately 30% of patients, and corticosteroid therapy is most often used.Until now there are no widely accepted standards of diagnosis of sarcoidosis. Commonly used techniques include: EBB, TBLB, TBNA, and mediastinoscopy. The novel diagnostic modality – endosonography (EBUS-NA and EUS-NA), introduced recently into clinical practice, is a promising alternative for mediastinoscopy, which until recently was the standard method of obtaining tissue samples from mediastinal lymph nodes. These methods were not verified in randomized studies.The role of thoracic surgeon in diagnosis of sarcoidosis is limited to invasive diagnostic procedures in cases, which could not be diagnosed using minimally invasive procedures. Such patients usually undergo mediastinoscopy or, rarely, VATS (in the stage III and IV of sarcoidosis). Thanks to the introduction of endoscopic procedures and ultrasound-guided biopsy, many patients may avoid surgery, i.e. mediastinoscopy.EBUS-NA and EUS-NA in the diagnostics of sarcoidosis were introduced in the Department of Thoracic Surgery of John Paul IInd Hospital in Cracow in 2008. Since then the number of surgeries (mediastinoscopy and VATS) in diagnostics of sarcoidosis decreased significantly. In order to compare methods used in diagnostics of sarcoidosis a prospective randomized study was designed. The aim of this thesis is a comparison of the standard endoscopic methods, such as EBB, TBLB and TBNA, with endosonographic methods: EBUS-NA and EUS-NA. Diagnostic yield of both endosonographic methods in the diagnosis of sarcoidosis was also compared. Moreover, the role of TBNA in diagnosis of sarcoidosis was assessed as a technically easier and cheaper method, which is not widely used.39In the period from October 2009 to May 2012, 106 patients were enrolled in the study. In all patients standard endoscopic methods were used including TBNA, EBB and TBLB and subsequently, after randomization, 55 patients had EBUS-NA and 51 had EUS-NA.All patients with negative results of the standard procedures and EBUS-NA underwent EUS-NA (8 patients) and all ; patients with negative results of the standard procedures and EUS-NA underwent EBUS-NA (8 patients). After additional EUS-NA 7 patients were diagnosed as having sarcoidosis, and after additional EBUS-NA 2 patients were diagnosed as having sarcoidosis. 4 patients with negative endoscopic procedures underwent mediastinoscopy. 6 patients met exclusion criteria.100 patients were included into the analysis. In all patients cyto-histopathological diagnosis of sarcoidosis was established. Using all minimally invasive methods (EBB, TBLB, TBNA as well as EUS-NA and EBUS-NA) the diagnosis was established in 96 patients (96%). Using standard endoscopic methods (EBB, TBLB and TBNA) the diagnosis was established in 64 patients (64%). Sensitivity and accuracy of standard endoscopic methods were both 64%. When analyzing each of the standard endoscopic methods separately, diagnosis was confirmed with EBB in 12 patients (12%), TBLB in 42 patients (42%), TBNA in 44 patients (44%). Using only EBB and TBLB the diagnosis was confirmed in 44 patients (44%). Additional use of TBNA increased sensitivity by 20%, which weighs in favour of performing TBNA in the early phase of diagnostics. Using only endosonographic methods (EBUS-NA and EUS-NA) the diagnosis was confirmed in 80 patients (80%). After comparing EBB+TBLB with each endosonographic method, a highly significant difference in sensitivity and accuracy was found in favour of endosonographic methods. The difference between accuracy and sensitivity of EBB+TBLB+TBNA and each endosonography method (p = 0.0112 and 0.0134) is statisti-cally significant in favour of these latter ones. Comparison of both endosonographic methods proved higher sensitivity and accuracy of EUS-NA, but the difference was not significant (p = 0.0724 and 0.0810). Thanks to additional biopsies 9 patients avoided invasive diagnostic procedures. None of the patients had tuberculosis confirmed in BACTEC of bronchial lavage and biopsy of lymph nodes.Conclusions:1. The diagnostic sensitivity and accuracy of endosonographic methods (EBUS-NA and EUS-NA) are significantly higher in comparison with the standard endoscopic methods (p < 0,01).2. Endosonographic techniques seem to be methods of choice in patients suspected of having sarcoidosis in the stage I and II.3. The diagnostic sensitivity and accuracy of EUS-NA is higher then EBUS-NA but the difference is not statistically significant.4. Combination of standard endoscopic methods, especially TBNA and TBLB, increase the yield of confirmation of sarcoidosis.