In order to determine whether or not irritation of the nasal passage with commonly used volatile anesthetics can elicit airway reflexes, we investigated respiratory, laryngeal, and tracheal responses to nasal insufflation of three volatile anesthetics (enflurane, isoflurane, and halothane) in 13 patients anesthetized with flunitrazepam, pentazocine, and nitrous oxide. The trachea of each patient was intubated with a saline-filled double-cuffed endotracheal tube. Changes in breathing pattern were measured with a pneumotachograph while changes in laryngeal wall tension and tracheal wall tension were assessed by measuring changes in the proximal cuff pressure and the distal cuff pressure, respectively. In 8 of 13 patients, the dose-response relationship for each anesthetic was determined by administering different concentrations (1, 3, and 5%) of gas mixtures. In these patients, nasal insufflation of 1 and 3% of each anesthetic did not produce any reflex response, whereas reflex responses were evident during nasal insufflation of 5% enflurane, isoflurane, and halothane. In all 13 patients, nasal insufflation of all three anesthetics at a concentration of 5% invariably produced changes in breathing pattern characterized by prolongation of expiratory time (TE). However, prolongation of TE was the most pronounced for enflurane (from a control value of 2.1 +/- 0.5 to a maximum value of 4.8 +/- 2.2 s [mean +/- standard deviation]), less for isoflurane (from 2.2 +/- 0.5 to 3.9 +/- 1.7 s), and the least for halothane (from 2.2 +/- 0.6 to 2.9 +/- 0.9 s).(ABSTRACT TRUNCATED AT 250 WORDS)