Several investigators have implicated norepinephrine and other toxic substances released in the region of a spinal cord injury in the genesis of the progressive pathological and clinical changes that follow spinal trauma. To test this hypothesis. we subjected cats to T-10 to T-12 laminectomy and monitored epidural spinal evoked potentials from sciatic nerve stimulation. The spinal subarachnoid space was perfused with normal saline, with norepinephrine solution, or with heparinized autologous blood or the pial surface of the spinal cord was exposed to macerated gray matter taken from the upper cervical cord. During 1- to 2-hour exposure periods, we noted no significant changes in the base line spinal evoked potentials. In another series of cats, we have shown that norepinephrine perfused over the spinal cord in this manner diffuses rapidly into the subpial white matter. Therefore, its failure to affect spinal evoked potentials does not represent a failure to penetrate the spinal cord. Putative toxins must originate either in extravasated blood or damaged neural tissue in the region of the spinal cord injury. The failure of ascending spinal tracts to react to blood or cord tissue in our experiment suggests that toxins are not involved in the spinal cord dysfunction that occurs soon after injury.