Traditional conscious sedation for endobronchial ultrasound procedures places patients at risk of desaturation, and high‐flow nasal oxygen may reduce the risk. We designed a parallel‐group randomised controlled trial of high‐flow nasal oxygen at a flow rate of 30–70 l.min−1via nasal cannulae, compared with standard oxygen therapy at 10 l.min−1 via a bite block in adults planned for conscious sedation for endobronchial ultrasound. The primary outcome was the proportion of patients experiencing desaturation (defined as SpO2 < 90%). Secondary outcomes included oxygen saturation after pre‐oxygenation, lowest oxygen saturation during procedure, number of hypoxic episodes, duration of hypoxia, end‐procedure end‐tidal CO2, satisfaction scores and complications. Thirty participants were allocated to each group. Baseline patient characteristics, procedure time and anaesthetic agents used were similar between the groups. Desaturation occurred in 4 out of 30 patients allocated to the high‐flow nasal oxygen group, compared with 10 out of 30 allocated to the standard oxygenation group, a non‐significant difference (p = 0.07) with intention to treat analysis. The difference was significant (p = 0.047) when using a per‐protocol analysis. Oxygen saturation after pre‐oxygenation and the lowest oxygen saturation during procedure were significantly higher in the high‐flow nasal oxygen group compared with the standard oxygenation group; median (IQR [range] 100 (99–100 [93–100]) vs. 98 (97–99 [94–100]), p = 0.0001 and 97.5 (94–99 [77–100]) vs. 92 (88–95 [79–98]), p < 0.001, respectively. There were no differences in other secondary outcomes. Although high‐flow nasal oxygen may prevent desaturation due to some causes, it does not protect against hypoxaemia in all circumstances.