In acute respiratory distress syndrome (ARDS), recruitment sessions of highfrequency oscillation (HFO) and tracheal gas insufflation (TGI) with short-lasting recruitment manoeuvres (RMs) may improve oxygenation and enable reduction of subsequent conventional mechanical ventilation (CMV) pressures. We determined the effect of adding HFO-TGI sessions to lung-protective CMV on early/severe ARDS outcome.We conducted a prospective clinical trial, subdivided into a first single-centre period and a second two-centre period. We enrolled 125 (first period, n554) patients with arterial oxygen tension (Pa,O 2 )/inspiratory oxygen fraction (FI,O 2 ) of ,150 mmHg for .12 consecutive hours at an end-expiratory pressure of o8 cmH 2 O. Patients were randomly assigned to an HFO-TGI group (receiving HFO-TGI sessions with RMs, interspersed with lung-protective CMV; n561) or CMV group (receiving lung-protective CMV and RMs; n564). The primary outcome was survival to hospital discharge.Pre-enrolment ventilation duration was variable. During days 1-10 post-randomisation, Pa,O 2 / FI,O 2 , oxygenation index, plateau pressure and respiratory compliance were improved in the HFO-TGI group versus the CMV group (p,0.001 for group6time). Within days 1-60, the HFO-TGI group had more ventilator-free days versus the CMV group (median (interquartile range) 31.0 (0.0-42.0) versus 0.0 (0.0-23.0) days; p,0.001), and more days without respiratory, circulatory, renal, coagulation and liver failure (pf0.003). Survival to hospital discharge was higher in the HFO-TGI group versus the CMV group (38 (62.3%) out of 61 versus 23 (35.9%) out of 64 subjects; p50.004).Intermittent recruitment with HFO-TGI and RMs may improve survival in early/severe ARDS.