Objective To explore the factors that are closely related to the sensitivity of endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) by analyzing the procedure-related factors, hoping to improve the operation of EBUS-TBNA. Methods A retrospective analysis was performed on all the 393 cases that had received EBUS-TBNA from Jan, 2015 to Dec, 2015 in Changhai Hospital, Second Military Medical University. Positive rates under different conditions such as working experiences of bronchoscopy operators, the types of patients, the types of anesthesia, the positions and sizes of biopsied lymph node, the numbers of biopsied lymph node groups, and the puncture times of each single lymph node were investigated. Logistic regression models and χ²-test were used to identify factors associated with the sensitivity of EBUS-TBNA. Results Our findings showed that the positive rates of EBUS-TBNA were not related to the different working periods as a bronchoscopy operators, types of patients (outpatient or hospitalization), types of anesthesia (local anesthesia or additional intravenous sedation to a local anesthetic), different positions of biopsied lymph node, the puncture times of each single lymph node or the numbers of different groups of lymph node chosen for biopsy. The maximum diameters of punctured lymph nodes under ultrasound interface were found related to the diagnosis rate of EBUS-TBNA (P<0.000 1). Among them, the positive rate of lymph node biopsy was significantly lower in those with the maximum diameter ≤1.0 cm than those with the maximum diameter>2.0 cm (P<0.000 1). The result showed that the positive rate of EBUS-TBNA was positively correlated with the size of biopsied lymph node, with a regression coefficient of 1.027(P<0.001). Conclusion Respiratory physicians can achieve satisfactory performance in EBUS-TBNA operation after formal training and extensive practice. In order to increase the positive rate and to improve the sensitivity of EBUS-TBNA, maximum diameter>2.0 cm lymph nodes should be chosen when possible. The numbers of biopsied lymph node groups, the puncture times of each single lymph node, the positions of biopsied lymph node, the type of anesthesia or the types of patients are not related to the positive rate of EBUS-TBNA.