It has been shown in mechanically ventilated awake normal humans that increasing inspiratory flow rate (VI) exerts an excitatory effect on respiratory output. It is not known if this effect persists during sleep. To test this, seven normal adults were studied during wakefulness and non-rapid eye movement (non-REM) sleep. Subjects were connected through a nose mask to a volume-cycled ventilator in the assist/control mode, and VI was increased in steps (3 to 4 breaths each) from 30 to 70 L/min and then back to 30 L/min. VI pattern was square, and all breaths were subject-triggered. Forty-one trials during non-REM sleep and 10 during wakefulness were analyzed. Both during sleep and wakefulness minute ventilation increased and total breath duration (Ttot) decreased significantly in a graded and reversible manner as VI increased. These changes were complete in the first breath after VI transition. The response was significantly less during sleep than during wakefulness (p < 0.050; at 30 L/min Ttot, expressed as percent of that at 70 L/min, was 110.2 +/- 1.3% during sleep and 127.8 +/- 3.9% during wakefulness. During wakefulness, the rate of change in airway pressure before triggering the ventilator (dp/dt), an index of respiratory drive, increased significantly (p < 0.05) with increasing VI. During sleep dp/dt was not affected by VI changes. In four sleeping subjects the increase in VI was sustained for 1.5 to 2 min. There was no evidence for adaptation of the response; Ttot, averaged over the last three breaths, did not differ from that obtained with VI was sustained for only 3 to 4 breaths. We concluded that VI exerts an excitatory effect on respiratory output, mediated by a reflex neural mechanism, and the gain of this reflex is attenuated by sleep.