1. Using both rectangular- and ramp-shaped intracellularly injected currents, the repetitive firing properties of 23 antidromically identified phrenic motoneurons were determined in anesthetized, paralyzed, and artificially ventilated cats during hypocapnic apnea. In response to rectangular depolarizing current injections, regular repetitive firing was observed in all cells. 2. At the beginning of a rectangular current pulse, the firing pattern was characterized by high frequency of firing that rapidly adapted to a much lower steady-state value. The relationship between the reciprocal of the first interspike interval (F1-2) and injected current was described by an initial linear portion of shallow slope, followed by a much steeper segment that smoothly reached a plateau value. The plateau value of F1-2 did not change with further increase in injected current. 3. The steady-state firing frequency versus injected current relationship was represented by a line of shallow slope over the entire range of injected currents. The slope of this line ranged between 1.1 and 4.5 Hz/nA. 4. A weaker correlation between minimal firing frequency for continuous activity and the reciprocal of the after hyperpolarization duration (1/AHPdur) was found for phrenic motoneurons than exists for lumbosacral motoneurons (26). Comparison of AHP shape at different levels of repetitive firing revealed that the slopes of the ascending portions of the AHP were similar except at the higher injected currents. Further, in the same cells during natural inspiratory activity the ascending part of the AHP was similar to that observed during current injection. 5. Depolarizing current ramps (approximately 1-s duration) were injected into 11 phrenic motoneurons. Instantaneous firing frequency was directly proportional to the intensity of the instantaneous injected current and independent of the rate of change of current for the range of ramp slopes tested (5-80 nA/s). Ramp-and-hold current injections were done in three motoneurons, and the peak instantaneous firing frequency showed little adaptation during the hold maneuver. 6. During hypocapnic apnea, we mimicked the expiratory-phase inhibition and inspiratory-phase excitation of phrenic motoneurons by injecting a 1-s depolarizing current ramp that was immediately preceded by a 1-s hyperpolarizing current ramp of the same absolute peak current intensity. Compared with the effects of positive current ramps alone the spike onsets during the negative-positive current ramp paradigm were either facilitated or retarded. Various ionic mechanisms are suggested for these effects as well as their function in determining the onset of firing during natur