Recording myoelectric motor-evoked potentials is frequently used as an in vivo evaluation technique in experimental studies of spinal cord injury (SCI). The aim of the present study was to determine whether specific neuronal pathways conduct these potentials. Stainless steel screws were permanently implanted into the cranium of 18 rats for stimulation of brainstem-evoked muscle potentials (B-MPs). Twelve rats were subjected to spinal cord lesions that restricted the continuity of the spinal cord to different discrete sections of the lateral and/or ventral white matter (WM) of the left hemicord. Sham rats (n = 6) were subjected to laminectomy only. Left hind limb B-MPs and motor function (open field walking test) were recorded before surgery and weekly thereafter for six consecutive weeks. Motor function was severely affected by SCI in all rats but recovered significantly during the first 14 postoperative days. The degree of functional recovery depended not only on the amount of spared WM but also on the particular section of WM that had been spared. In contrast, B-MP amplitudes also were severely reduced by SCI, but did not recover during the survival period. Moreover, B-MP amplitudes correlated only weakly with the amount of sparedWM and were not influenced by which section ofWM had been spared. While functional recovery correlated significantly with the amount of spared WM, no correlation was found between B-MP amplitudes and functional recovery. B-MP conduction velocities were not affected by SCI. It is therefore believed that B-MPs have little prognostic value for experimental studies of SCI in the rat.