Endobronchial ultrasound (EBUS)-transbronchial needle aspiration (TBNA) has become a widely available tool that allows sampling of mediastinal and hilar lymph nodes with comparable accuracy as compared with the gold standard procedure, mediastinoscopy. The goal of this study was to evaluate the competence accuracy of this technique in academic medical center in patients with and without malignant disease. This was a retrospective chart review of the first 150 patients who underwent EBUS-TBNA at our institution with an operator not previously trained or supervised while performing the procedure. All nondiagnostic results were confirmed with mediastinoscopy. The cumulative sum analysis is a method used to continuously monitor performance against an established standard to attain competence in the procedure performed. Learning curve was assessed using cumulative sum method. Procedures were divided into sextiles (1–25, 26–50, 51–75, 76–100, 101–125, and 126–150). The technique's diagnostic accuracy was calculated for each of the 6 categories and trend toward improved accuracy was assessed using Cochran–Armitage trend test. Operator competency was achieved between 55th and 60th procedures. The diagnostic accuracy increased from 72% to 88% (from the first to third sextile) but remained stable afterwards at 88% (C-A trend test P = 0.091). The overall diagnostic accuracy was 84%. Trainees' learning rate varies while acquiring adequate knowledge. We suggest that a learning curve for each operator be used to assess competence in EBUS-TBNA procedure before physicians perform it without supervision.