Corticomotor evoked potentials have recently been used in experimental animals and patients as a measure of neurologic function after stroke. However, little is known about the fundamental electrophysiologic properties contributing to the formation of these potentials. To define some of these properties, corticomotor evoked potentials were recorded from the contralateral hindlimb in response to transcortical stimulation in cats anesthetized with halothane. These potentials were obtained hourly for 6 hours after middle cerebral artery occlusion. Four major identifiable components were recorded in control responses. Immediately after infarction, all component amplitudes were significantly attenuated. However, after approximately 5 hours, the early latency components exceeded control values; late latency components were also increased. Corresponding somatosensory evoked potentials were abolished and did not return throughout the recording session. Based on classic neurophysiologic studies, the amplitude increment can be explained as combined activation of low-threshold brainstem facilitatory centers and/or direct activation of subcortical axonal pathways. With further study, corticomotor evoked potentials may be a valuable adjunct to current electrophysiologic monitoring techniques, particularly with regard to defining the extent and location of an infarct, as well as to assessing functional recovery. (Stroke 1987;18:1141-1147) S omatosensory evoked potentials (SSEPs) have been used to assess neurologic function for decades. 1 ' 2 Both normal and pathologic conditions of the central nervous system have been studied by SSEP analysis in humans and in many animal models.34 Such investigations have historically included SSEP evaluation in cerebral ischemia.3 " 7 Several attempts have been made to correlate cerebral blood flow, infarct size, or neurologic deficit to changes in SSEP waveforms. 89 SSEPs, however, primarily reflect the integrity of the sensory system. 10 " 12 Extrapolation of motor function from altered features in SSEP characteristics has been criticized. 1213 Residual motor function following a cerebral ischemic event is of considerable therapeutic and prognostic importance. Therefore, a noninvasive, quantitative, and objective tool for assessing postinfarct motor function is desirable.Recently, a transcranial method for noninvasive stimulation of the motor cortex has been developed.1413 By this method, motor evoked potentials produced by the spinal cord, peripheral nerves, or muscles can be recorded in response to transcranial stimulation. Very few studies have employed corticomotor evoked Received April 4, 1987; accepted June 24, 1987. potentials (CMEPs) as a technique for evaluating stroke."-17 Although CMEPs should prove to be a valuable adjunct to methods of patient assessment, relatively few principles of normal waveform formation or component alterations secondary to injury have been clearly defined. The present study was undertaken in an attempt to clarify the effect of acute stroke in CMEP ...