Upper airway resistance (UAR) increases in normal subjects during the transition from wakefulness to sleep. To examine the influence of sleep on upper airway collapsibility, inspiratory UAR (epiglottis to nares) and genioglossus electromyogram (EMG) were measured in six healthy men before and during inspiratory resistive loading. UAR increased significantly (P less than 0.05) from wakefulness to non-rapid-eye-movement (NREM) sleep [3.1 +/- 0.4 to 11.7 +/- 3.5 (SE) cmH2O.1-1.s]. Resistive load application during wakefulness produced small increments in UAR. However, during NREM sleep, UAR increased dramatically with loading in four subjects although two subjects demonstrated little change. This increment in UAR from wakefulness to sleep correlated closely with the rise in UAR during loading while asleep (e.g., load 12: r = 0.90, P less than 0.05), indicating consistent upper airway behavior during sleep. On the other hand, no measurement of upper airway behavior during wakefulness was predictive of events during sleep. Although the influence of sleep on the EMG was difficult to assess, peak inspiratory genioglossus EMG clearly increased (P less than 0.05) after load application during NREM sleep. Finally, minute ventilation fell significantly from wakefulness values during NREM sleep, with the largest decrement in sleeping minute ventilation occurring in those subjects having the greatest awake-to-sleep increment in UAR (r = -0.88, P less than 0.05). We conclude that there is marked variability among normal men in upper airway collapsibility during sleep.