BackgroundThe bronchial suction has been applied in speeding lung collapse. Low suction pressure may not speed lung collapse, but high pressure causes occult lung injury. The aim of the study was to explore efficacy and safety of different suction pressure for speeding lung collapse.MethodsEighty-four subjects undergoing uniport video-assisted thoracoscopic surgery (VATS) were randomly assigned for non-suction (Group 0), -10 cmH2O suction pressure (Group − 10), and − 30 cmH2O suction pressure (Group − 30). The primary outcome were the lung collapse scores (LCS) at 0 min (T0) after the visualization of the lung using a 10-point visual analogue scale and area under the curve (AUC) of LCS over time. The secondary outcomes included disconnection from the ventilator, the assessment of occult lung injury using NOS-3 expression, histologic scores of lung injury, and lung W/D weight ratio, intraoperative hypoxemia, the incidence of perioperative pulmonary complications.ResultsBoth the LCS at T0 and AUC analysis showed that compared with Group 0, Group − 10 and Group − 30 significantly achieved good lung collapse (P < 0.05), but no difference between Group − 10 and Group − 30. Four patients in Group 0 were treated with disconnection maneuver. The assessment of occult lung injury showed no differences.ConclusionsApplying − 10 cmH2O suction pressure for 1 min when pleural incision is a relatively safe method to promote lung collapse without the occurrence of occult lung injury.Trial registrationChinese Clinical Trial Registry number, ChiCTR2200062991. Registered on 26/08/2022.