The purpose of the present study was to compare in awake and asleep healthy subjects, under nasal intermittent positive pressure ventilation (nIPPV) with a two-level intermittent positive pressure device (two-level nIPPV), the efficacy of the controlled and spontaneous modes, and of different ventilator settings in increasing effective minute ventilation (VE). Eight healthy subjects were studied. In the controlled mode, inspiratory positive airway pressure (IPAP) was kept at 15 cm H2O, expiratory positive airway pressure (EPAP) at 4 cm H2O, and the inspiratory/expiratory (I/E) time ratio at 1. The respirator frequencies were 17 and 25/min. In the spontaneous mode experiment, IPAP was started at 10 cm H2O and progressively increased to 15 and 20 cm H2O; EPAP was kept at 4 cm H2O. We measured breath by breath the effective tidal volume (VT with respiratory inductive plethysmography), actual respiratory frequency (f), and effective VE. Using the controlled mode, effective VE was significantly higher on nIPPV than during spontaneous unassisted breathing, except in stage 2 nonrapid eye movement sleep at 17/min of frequency; increases in f from 17 to 25/min led to a significant decrease in VT reaching the lungs, during wakefulness and sleep; effective VE was higher at 25 than at 17/min of frequency only during sleep; periodic breathing was scarce and apneas were never observed. Using the spontaneous mode, with respect to awake spontaneous unassisted breathing, two-level nIPPV at 10 and 15 cm H2O of IPAP did not result in any significant increase in effective VE either in wakefulness or in sleep; only IPAP levels of 20 cm H2O resulted in a significant increase in effective VE; during sleep, effective VE was significantly lower than during wakefulness; respiratory rhythm instability (ie, periodic breathing and central apneas) were exceedingly common, and in some subjects extremely frequent, leading to surprisingly large falls in arterial oxygen saturation. It appears that two-level nIPPV should be used in the controlled mode rather than in the spontaneous mode, since it seems easier to increase effective VE with a lower IPAP at a high frequency than at a high pressure using the spontaneous mode. We suggest that the initial respirator settings in the controlled mode should be an f around 20/min, an I/E ratio of 1, 15 cm H2O of IPAP, and EPAP as low as possible.