Background/aims: Endoscopic ultrasound-guided fine needle aspiration is an established tissue-acquisition technique for mediastinal lesions. However, there are limitations to endoscopic ultrasound-guided fine needle aspiration of mediastinal masses in certain neoplasms and granulomatous diseases. Most studies have used 22-gauge aspiration and/or 19-gauge Tru-cut needles, and only limited data exist on larger-caliber aspiration needles. We aimed to compare prospectively the diagnostic yield of endoscopic ultrasound-guided fine needle aspiration using 19- and 22-gauge aspiration needles in patients with mediastinal lesions of unknown origin. Material and Methods: Using a consecutive entry design, 57 patients with mediastinal mass or lymph node, in whom previous investigations, including bronchoscopy and computed tomography-guided biopsy, had not provided a final diagnosis, underwent endoscopic ultrasound-guided fine needle aspiration biopsy using 19-gauge or 22-gauge aspiration needle. Determination of the adequacy and cytopathologic interpretation of fine needle aspiration materials were done by two pathologists blinded to the clinical condition of the patient. Fine needle aspiration specimens were placed in four categories as: (1) nondiagnostic, (2) benign, (3) granulomatous disease, and (4) malignant. Results: Among 57 patients [35 (61.4%) with mediastinal lymph nodes and 22 (38.5%) with pulmonary masses], adequate tissue was obtained in 52 (91.2%) of the cases (with a mean of 3.3 needle passes). Correct cytopathologic diagnoses were made based on the endoscopic ultrasound-guided fine needle aspiration specimens obtained by 19- and 22-gauge needles in 96% and 92% of the samples, respectively (p>0.05). Conclusions: As concerns endoscopic ultrasound-guided fine needle aspiration of mediastinal masses and lymph nodes, the diagnostic sensitivity of aspirated material obtained using 19- and 22-gauge fine needle aspiration needles was found to be comparable in our study.