1. The patterns of membrane potential changes of phrenic motoneurons were compared during fictive vomiting, fictive coughing, and fictive swallowing in decerebrate, paralyzed cats. These fictive behaviors were identified by motor nerve discharge patterns similar to those recorded from the muscles of nonparalyzed animals. Phrenic motoneurons (n = 54) were identified by antidromic activation from the thoracic phrenic nerve. Intracellular recordings were obtained from 27 motoneurons during fictive vomiting, 40 during fictive coughing, and 27 during fictive swallowing. Sixteen motoneurons were recorded during both fictive coughing and fictive swallowing, eight during both fictive coughing and fictive vomiting, and two during both fictive vomiting and fictive swallowing. Seven motoneurons were studied during all three behaviors. 2. Fictive vomiting, typically evoked by electrical stimulation of abdominal vagal afferents, was characterized by a series of bursts of coactivation of phrenic and abdominal motor nerves, culminating in an expulsion phase in which abdominal discharge was prolonged both with respect to phrenic discharge and to abdominal discharge during the preceding retching phase. During fictive vomiting, phrenic motoneurons depolarized abruptly, and the amplitude of depolarization was significantly greater than during control inspirations. They then repolarized slowly throughout the phrenic burst, rapidly repolarizing at the end of each phrenic burst during retching and reaching a level similar to that observed during expiration. During the expulsion phase, the pattern was initially the same. However, after the cessation of phrenic discharge, the membrane potential repolarized slowly until the end of the abdominal burst, exhibiting greater synaptic noise than during expiration. One phrenic motoneuron, presumably innervating the periesophageal region of the diaphragm, received a strong hyperpolarization just before the onset of the emetic episode and fired for shorter periods during fictive vomiting than did other phrenic motoneurons. Reversal of inhibitory postsynaptic potentials (IPSPs) by chloride ion and/or current injections into six motoneurons revealed the presence of inhibition during the period between phrenic bursts during fictive vomiting and also during the final phase of expulsion when phrenic discharge ceased but abdominal discharge continued. 3. Fictive coughing, evoked by repetitive electrical stimulation of superior laryngeal nerve afferents, was characterized by a large phrenic discharge followed immediately by a large abdominal nerve discharge. During fictive coughing, phrenic motoneurons retained their ramplike depolarizations throughout phrenic discharge; however, the amplitude of depolarization was greater than during inspiration. During the subsequent abdominal nerve discharge, the phrenic membrane potential usually underwent an initial rapid, transient hyperpolarization followed by a gradual repolarization associated with increased synaptic noise. Reversal of IPSPs in one motoneuron revealed a strong inhibition during the period of abdominal nerve discharge. 4. The buccopharyngeal stage of fictive swallowing was characterized by brief bursts of activity in the pharyngeal vagus and hypoglossal nerves and was usually observed both after and episode of fictive vomiting and in response to repetitive electrical stimulation of superior laryngeal afferents. During fictive swallowing, phrenic motoneurons exhibited a brief depolarization that was half the amplitude of that observed during inspiration. No IPSPs were observed either before, during, or after the depolarization. 5. In conclusion, the descending central drives to phrenic motoneurons, as revealed by the patterns of membrane potential changes, differ markedly during fictive vomiting, coughing, and swallowing. In addition to excitation, inhibitory inputs are involved in shaping phrenic motoneuronal discharge during vomiting and coughing.