Oscillatory ventilatory pattern occurs more frequently in sleep despite the stabilizing factor of sleep-induced reduction in CO2 chemosensitivity. In nine young normal humans, we have tested the hypothesis that, despite a sleep-induced reduction in chemosensitivity, the transient central chemoreceptor-mediated change inspiratory ventilation (VI) caused by a standardized disturbance to chemoreflex ventilatory control is similar in quiet sleep and wakefulness. The equivalent VI response to a single-breath hyperoxic hypercapnic stimulus (i.e., inhaling a single breath of 0.01 liter of CO2 in O2--a direct measure of "closed-loop" dynamic response) was determined using pseudorandom binary CO2 stimulation and the prediction-error method of transfer function estimation. From these data, the response of VI to a single-breath increase of 1 Torr in end-tidal PCO2 was also derived, from which "dynamic" central chemosensitivity was calculated. Despite a 43% reduction in dynamic central chemosensitivity, the peak and the area under the closed-loop VI response are similar in wakefulness and quiet sleep, whereas sleep increases the duration of the response by 48%. Thus hyperoxic ventilatory stability is not reduced in quiet sleep relative to wakefulness. We propose that changes in dynamics of pulmonary gas exchange in sleep substantially offset the decreased chemosensitivity, thereby maintaining the gains and time constants of the central chemoreceptor-mediated component of the closed-loop ventilatory control system similar to those during wakefulness.