The movements produced by submaximal tetanic stimulation of the medullary pyramid in the decerebrate cat have been studied by multiple electromyographic recording. A variety of types of response ranging from rather steady posture holding to regular phasic “walking” movements was found in different preparations and in the same preparations at different times. Not all of the causes for variation could be defined experimentally. Inhibition was found commonly to be a component of pyramidal response, and was not related invariably to reciprocal contraction. Definable factors which could influence grossly the overall form, degree, direction and distribution of muscular contraction were stimulus intensity, posture, afferent nerve stimulation, and collateral pyramidal stimulation. There was no evidence for a proximodistal muscle threshold relationship. Tibialis anticus in the hindlimb usually responded most readily, but exceptions were numerous. Ipsilateral movements were usually slight in degree and appeared to be postural reactions. No evidence of a significant uncrossed pathway could be demonstrated. Forelimb and hindlimb were equally represented in all cross-sectional areas of the pyramids. Variation of stimulus parameters within wide limits did not affect the response pattern; threshold was to an approximation a function of stimulating power. It is inferred from the results that a large degree of the variation of response to cortical or pyramidal stimulation is a result of the complexities of integrative channeling by the spinal mechanism and that Jackson's second level of motor function must include this internuncial mechanism. It is also concluded that pyramidal function has two related aspects, a general facilitatory action on spinal internuncial patterns, and a more specific directive effect of the pattern of pyramidal activity on the spinal mechanism. Arbitrary pyramidal stimulation chiefly manifests the former. In view of these findings, the validity of inferring detailed functional and anatomical relation from the response to artificial stimulation of the CNS above the spinal level is considered and it is concluded that such techniques alone are inadequate for defining physiological mechanism. It is also suggested that physiological patterns of total neural activity may be of more critical importance in the organization of movement than are structural relations alone.