OBJECTIVE Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has typically been performed using the 22 gauge (G) dedicated TBNA needle. Recently a new 21G TBNA needle has been introduced. The efficacy of using a larger gauge biopsy needle during EBUS-TBNA has not been reported. The purpose of this study was to compare the diagnostic yield and utility of 21G and 22G needles during EBUS-TBNA. METHODS EBUS-TBNA was performed using both 21G and 22G needles. Cytological and histological findings were recorded for each samples obtained by an independent cytologist and pathologist. The cellularity and blood contamination were evaluated in the cytological samples. The quality of the histological core was evaluated by the amount of blood clots versus the actual tissue. Each factor was compared within two slides from the two different size needles. The diagnostic yield and the differences of the cytology and histology were analysed. RESULTS The evaluation of 45 lesions by EBUS-TBNA revealed that tumour cells were equally detected by both 21G and 22G needles. Two patients of adenocarcinoma were histologically diagnosed only by the 21G needle. Although histological structure was better preserved in five lesions collected by the 21G needle, there was more blood contamination with the 21G needle (P < 0.0001). CONCLUSIONS There were no differences in the diagnostic yield between the 21G and 22G needles during EBUS-TBNA. The preserved histological structure of the samples obtained by the 21G needle may be useful for the diagnosis of mediastinal and hilar adenopathy of unknown aetiology which may be a challenge with the 22G needle.