The nasal airway is a common route for oxygen delivery to hypoxemic patients, and it has been advocated as a suitable route for intermittent positive pressure ventilation to patients in respiratory failure. There is, however, conflicting evidence on the question of whether nasal airflow has a stimulant or a depressant effect on ventilation. We therefore studied the effects on ventilation of increasing and decreasing nasal airflow during sleep, thereby avoiding the voluntary and behavioral influences on breathing seen during wakefulness. After an acclimatization night, each of nine normal male volunteers (20 to 28 yr of age) underwent overnight sleep studies using standard techniques. Each sleep study had three phases of at least 2-h durations: (1) control nasal breathing (CNB); (2) added nasal flow of 4 L of compressed air via nasal prongs (ANF); (3) predominant mouth breathing (PMB). Significant differences were found by ANOVA for the three experimental periods for both minute ventilation (Vl) and mean inspiratory flow rate (VT/Tl) during Stage 2 sleep, with the highest values occurring during ANF. Vl was 5.40 +/- 0.58 L/min (mean +/- SD) during ANF, 5.35 +/- 0.82 during CNB, and 4.92 +/- 0.71 during PMB (p < 0.05 by ANOVA). VT/Tl was 266 +/- 23 ml/s during ANF, 248 +/- 39 during CNB, and 241 +/- 30 during PMB (p < 0.02 by ANOVA). VT also tended to be higher during ANF (p < 0.1 by ANOVA). Respiratory frequency did not differ significantly during the three study periods. There was insufficient slow-wave and REM sleep to allow ventilatory comparisons in these sleep stages.(ABSTRACT TRUNCATED AT 250 WORDS)