The activity of corticomotoneuronal (CM) cells, identified by clear postspike facilitation (PSF) of rectified EMG [electromyogram] activity in spike-triggered averages, was recorded in the precentral cortex of monkeys making 2 types of ramp-and-hold wrist responses. Auxotonic wrist movements against elastic loads required active torque proportional to wrist displacement, and isometric responses involved ram-and-hold torque trajectories with no wrist displacement. On the basis of their firing pattern during the ramp-and-hold responses, all of the 135 CM cells were classified into 4 types: phasic-tonic (59%), tonic (28%), phasic-ramp (8%), or ramp (5%). All CM cells were active during the static hold period; tonic cells discharged at a constant rate, while ramp cells showed steadily increasing discharge during the hold period. During the dynamic phase of the response, i.e., during the torque ramp, the phasic cells exhibited an additional peak of activity exceeding the final tonic level associated with the hold period. Single CM cells exhibited the same response pattern in association with both isometric and auxotonic responses; thus, the torque trajectory rather than displacement or velocity is apparently the primary determinant of the CM cell''s response pattern. Other precentral cells, which discharged phasically at onset of movement, but exhibited no tonic discharge during the hold period did not produce PSF. The 4 types of CM cells did not differ significantly in the size of their muscle field, their response to passive movements, or their location in the cortex. The onset time of CM cell activity relative to the onset of EMG activity in its facilitated target muscle(s) was measured from response averages. The phasic-tonic and phasic-ramp cells began firing significantly earlier with respect to their target muscles than the tonic and the ramp cells. Since the peak PSF occurred about 10 ms after the spike, the initial discharges in many CM cells would contribute to subthreshold facilitation of their target motoneurons. The relation between firing rate of CM cells and active torque was studied. Over the range of torque studied, only a few CM cells had appreciable nonzero torque thresholds for tonic firing. Most cells were inactive with the monkey at rest and during movement opposite their favored direction. Thus, most CM cells appear to be recruited at low torque levels; their contribution to increases in active torque is due more to increases in firing rate than to recruitment of additional CM cells with high thresholds. Some CM cells, whose activity covaried reliably with controlled ramp-and-hold responses, became less active during rapid ballistic responses, suggesting some differences in neural mechanisms underlying these 2 types of movement.