SummaryInternational recommendations encourage liberal administration of oxygen to patients having surgery under general anaesthesia, ostensibly to reduce surgical site infection. However, the optimal oxygen regimen to minimise postoperative complications and enhance recovery from surgery remains uncertain. The hospital operating theatre randomised oxygen (HOT‐ROX) trial is a multicentre, patient‐ and assessor‐blinded, parallel‐group, randomised clinical trial designed to assess the effect of a restricted, standard care, or liberal peri‐operative oxygen therapy regimen on days alive and at home after surgery in adults undergoing prolonged non‐cardiac surgery under general anaesthesia. Here, we report the findings of the internal vanguard feasibility phase of the trial undertaken in four large metropolitan hospitals in Australia and New Zealand that included the first 210 patients of a planned overall 2640 trial sample, with eight pre‐specified endpoints evaluating protocol implementation and safety. We screened a total of 956 participants between 1 September 2019 and 26 January 2021, with data from 210 participants included in the analysis. Median (IQR [range]) time‐weighted average intra‐operative F_iO₂was 0.30 (0.26–0.35 [0.20–0.59]) and 0.47 (0.44–0.51 [0.37–0.68]) for restricted and standard care, respectively (mean difference (95%CI) 0.17 (0.14–0.20), p < 0.001). Median time‐weighted average intra‐operative F_iO₂was 0.83 (0.80–0.85 [0.70–0.91]) for liberal oxygen therapy (mean difference (95%CI) compared with standard care 0.36 (0.33–0.39), p < 0.001). All feasibility endpoints were met. There were no significant patient adverse events. These data support the feasibility of proceeding with the HOT‐ROX trial without major protocol modifications.