1. The details of the organization of the motor cortex and its anterior and posterior border were investigated in three monkeys by a combination of techniques including intracortical microstimulation (i.c.m.s.), electrophysiological recording of cutaneous and muscle afferent inputs to single cortical neurones, and electrophysiological and anatomical identification of corticospinal neurones; in addition, data from these methods were related to cortical cytoarchitecture.2. Almost 5000 individual cortical loci were tested with i.c.m.s. in the unanaesthetized monkeys. In this paper, we particularly consider the organization of the forelimb motor representation, and its relation to the representation of other parts of the body. I.c.m.s. thresholds of about 5 muA were common for evoking twitch movements and e.m.g. responses in distal forelimb and face, jaw and tongue muscles, but proximal forelimb, trunk and hind-limb movements also sometimes had such low thresholds.3. The fingers were found to be represented nearest the central sulcus, with horseshoe-shaped bands of cortical tissue representing progressively more proximal muscles situated around this central ;finger core'.4. Cytoarchitectonically, the cortex having these low-threshold motor effects was characteristic of area 4. There was also a close fit between the extent of this ;excitable cortex' and the extent of densely spaced corticospinal neurones identified electro-physiologically or with horseradish peroxidase labelling. In subsequent mapping of forelimb afferents to the cortex when the animal was deeply anaesthetized, low-threshold and short-latency responses to muscle nerve stimulation were rarely found in this ;excitable cortex'.5. The anterior border could be clearly established by i.c.m.s. and by the sharp boundary of corticospinal neurones. It was noted that the motor cortex extends rostrally beyond area 4 and its anterior border appears to reside in the posterior part of area 6aalpha (Vogt & Vogt, 1919) although it is difficult to establish the precise transition from area 4 to area 6.6. Posteriorly, the ;micro-excitable cortex' was found to be limited to regions cytoarchitectonically delineated as area 4 and did not include area 3a. On the other hand, low-threshold forelimb proprioceptive afferent inputs appeared restricted to area 3a neurones in the deeply anaesthetized animal. Corticospinal neurones were very dense in area 4, and there was a clear decrease in their occurrence in more caudal areas. However, scattered nests of corticospinal neurones were noted in areas 3a, 3b, 2, 1 and 5. It remains to be seen whether these scattered nests could be directly involved in motor control or whether they may modulate ascending somatosensory transmission, and whether they rely on sensory feed-back or inputs from other central areas for their spinal effects.