BACKGROUND: An optimal sampling sequence in radial guide sheath endobronchial ultrasound lung biopsy (R-EBUS) is unclear. This prospective single-center pilot randomized controlled trial aimed to determine if the initial method and sequence of sampling affect the diagnostic accuracy of the procedure. METHODS: Consecutive patients undergoing R-EBUS for lesions >15 mm with a bronchus sign were randomly assigned (1:1:1) to biopsy first (group A), brushings first (group B) or combination (group C). The primary outcome was a positive diagnosis from any sampling method. RESULTS: Fifty-four patients were randomized. The overall diagnostic yield of the procedure was 77.8% (95% confidence interval: 66%-89%), with no difference between groups. A higher rate of positive cytology from brushings was seen if the biopsies were performed before brushings (77.8% in group A vs. 44.4% in group B, P=0.03). The rate of positive cytology from washings was higher if the washings were obtained just after the brushings (61.1% in group A vs. 11.1% in group B, P=0.02). There was no difference in the rate of positive biopsy histology in the groups (P=0.27). All 3 sampling modalities were more likely to be positive in group A (50.0% vs. 11.1% in group B and 22.2% in group C, P=0.04). Complications rate was low and not significantly different between groups. CONCLUSION: The overall rate of a positive R-EBUS procedure was not affected by the initial sampling method or sequence. However, all 3 sampling modalities were more likely to be positive if biopsies were performed first, followed by brushings and washings.