Transcranial magnetic stimulation (TMS) can activate the corticobulbar system and briefly recruit upper airway dilator muscles, improving the inspiratory airflow dynamics of flow-limited respiratory cycles during sleep. The purpose of this investigation was to quantify the effects of TMS-induced twitches applied during sleep on flow-limited respiratory cycles in 14 obstructive sleep apnoea patients. Submental muscle motor threshold (SUBMT) and motor-evoked potential (SUBMEP) were examined during wakefulness and sleep. The TMS-induced twitches were applied during stable non-rapid eye movement (NREM) sleep, during non-consecutive flow-limited respiratory cycles at the beginning of inspiration, with intensities varying from sleep SUBMT up to maximal stimulation without arousal. Maximal inspiratory flow, inspiratory volume, shifts of electroencephalogram frequency and pulse rate variability were assessed. Cortical and/or autonomic arousal after TMS was observed in only 13.8% of all twitches applied. The SUBMT increased during NREM sleep (wakefulness, 24.8 ± 9.3%; and NREM sleep, 28.3 ± 9.5%;P= 0.003). Augmenting stimulator output from SUBMT to maximal stimulation before arousal enhanced SUBMEP peak-to-peak amplitude (from 0.09 ± 0.05 to 0.4 ± 0.3 mV;P= 0.005) with a concomitant rise in maximal inspiratory flow (from 376.2 ± 107.9 to 411.9 ± 109.3 ml s−1;P= 0.008) and inspiratory volume (from 594.8 ± 189.2 to 663.7 ± 203.1 ml;P= 0.001) in all but one patient. Corticobulbar excitability of submental muscles decreases during NREM sleep. Brief recruitment of submental muscles with TMS during sleep improves upper airway mechanics without arousing patients from sleep.