1. Our purpose was to characterize the ventilatory patterns of eupnoea and gasping in the neonatal rat. This study was precipitated by reports, using in vitro brainstem spinal cord preparations, that only a single pattern is present in neonatal rats. 2. In anaesthetized or decerebrate rat pups aged less than 13 days, eupnoea was characterized by a sudden onset of inspiratory activity and then a more gradual rise to peak levels. Following vagotomy, frequency fell and peak phrenic activity and tidal volume increased. The rate of rise of inspiratory activity also rose, but peak levels were still achieved during the latter half of inspiration. Vagal efferent activity exhibited bursts during both inspiration and the early expiration. This basic eupnoeic rhythm was not altered after sectioning of the carotid sinus nerves. 3. Upon exposure to hypoxia or anoxia, phrenic activity, tidal volume and frequency initially increased and then declined. In many animals, ventilatory activity then ceased, but later returned with a gasping pattern. 4. Gasping was characterized by a sudden onset of phrenic activity, which reached a peak intensity during the early portion of inspiration. The expiratory burst of vagal activity was eliminated. 5. Reductions of body temperature from 37 to 27 degrees C resulted in prolongations of inspiration and expiration and decreases of phrenic amplitude; phasic phrenic activity completely disappeared in some animals. Upon exposure to anoxia, gasping was observed, even in animals in which phrenic activity had disappeared in hyperoxia. 6. We conclude that, from the day of birth, rats can exhibit eupnoea and gasping patterns which are very similar to those of adult animals. 7. The rhythmic neural activities of the in vitro brainstem-spinal cord preparation, reported by others, differ markedly from eupnoea but are identical with gasping. We therefore conclude that this preparation is not suitable for investigation of the mechanisms that generate eupnoeic breathing.