Inability to breathe through the nose is an increasingly recognized cause of disordered breathing during sleep. To test the hypothesis that this respiratory dysrhythmia could result from loss of neuronal input to respiration from receptors located in the nose, we anesthetized the nasal passages of 10 normal men during sleep. Each subject spent 4 consecutive nights in the sleep laboratory while sleep stages, breathing patterns, respiratory effort, and arterial oxygen saturation were monitored. Night 1 was for acclimatization with Nights 3 and 4 being randomized to nasal spraying with either 4% lidocaine or placebo. On the lidocaine and placebo nights (Nights 3 and 4) the nasal passages were also sprayed with a decongestant to prevent increased nasal air-flow resistance resulting from mucosal swelling. To control for the possible effects of this decongestant, an additional night (Night 2) was included during which the nasal passages were sprayed with room air. Parallel studies conducted during wakefulness demonstrated low nasal resistance during the lidocaine-decongestant regimen. Because of the short duration of anesthesia with lidocaine, spraying was done at lights out and 2.5 and 5 h later. On the placebo night (decongestant plus saline) there were 6.4 +/- 1.8 (SEM) disordered breathing events (apneas plus hypopneas) per subject, whereas with lidocaine (plus decongestant) this increased fourfold to 25.8 +/- 7.8 events per subject (p less than 0.05). The majority of the disordered breathing events were apneas and were fairly evenly distributed between central and obstructive events. The magnitude of these changes is similar to that previously reported with complete nasal obstruction. These results suggest that nasal receptors sensitive to air flow may be important in maintaining breathing rhythmicity during sleep.