1. Relationships between the timing of respiration and deglutition were studied in thirty awake healthy subjects at rest. Deglutition was monitored by submental electromyography, pharyngeal manometry and videofluoroscopy. Respiration was recorded by measurement of oronasal airflow and chest wall movement. Three types of deglutition were studied: injected bolus swallows, spontaneous swallows, and visually cued swallows of boluses previously placed in the mouth. 2. The effect of each swallow on respiratory rhythm was characterized by measurement of cophase, defined as the interval between the onset of deglutitive submental EMG activity to the onset of subsequent rescheduled inspirations. Cophase was determined for swallows initiated at different phases of the respiratory cycle. In all subjects deglutition caused phase resetting of respiratory rhythm. Cophase was largest for swallows initiated near the inspiratory-expiratory (I-E) transition and smallest for swallows initiated near the expiratory-inspiratory (E-I) transition. The pattern of respiratory resetting by deglutition was topologically classified as type 0. This pattern was shown for swallows induced by bolus injection or visual cue, and for spontaneous swallows. 3. The incidence of spontaneous deglutition was influenced by the position of the swallow in the respiratory cycle. Few spontaneous swallows were initiated near the E-I transition whereas most occurred from late inspiration to mid-expiration. 4. Deglutition caused an abrupt decrease in airflow leading to an interval of apnoea, followed by a period of expiration. The duration of deglutition apnoea for spontaneous swallows was shorter than that for 5 ml bolus swallows, and was unaffected by the respiratory phase of swallow initiation. The period of expiration after swallowing was longest for swallows initiated at the I-E transition, and shortest for E-I swallows. 5. The intervals between bolus injection and the onset of deglutition apnoea, and the timing of swallowing events, were not significantly altered by the phase in the respiratory cycle at which swallowing was exhibited. 6. To quantify the relationship between bolus flow and respiration, we determined the latencies between cessation of inspiratory airflow and arrival of the bolus at the larynx (a.), and between laryngeal bolus departure and resumption of inspiratory airflow (a). Both values were dependent upon the respiratory phase of swallowing. The lowest values for a and a were found for early-inspiratory and late-expiratory swallows, respectively. 7. We conclude that swallowing causes respiratory phase resetting with a pattern that is characteristic of the strong perturbations of an attractor-cycle oscillator. The threshold for initiation of swallowing in awake subjects is influenced by, but not strongly coupled to, the phase of respiration. We propose that respiratory timing, in addition to anatomical barriers within the upper airway, influences the vulnerability for aspiration during deglutition. Swallows initiated near the E-I transiti...