During tail regeneration most lizards also regenerate the tail spinal cord. The regenerated spinal cord primarily contains neuroepithelium (i.e., the ependymal tube which forms the central canal) and descending axons. The present experiments identify the source of the axons in the regenerated spinal cord. Application of HRP to normal tail spinal cord resulted in labeled cells in the nucleus paraventricularis, the interstitial nucleus of the fasciculus longitudinalis medialis, the nucleus ruber, the medullary reticular formation (including raphe nuclei), as well as in vestibular nuclei. HRP applied to the regenerated spinal cord labeled only 4% of the cells seen in normal animals, and these were confined to rhombencephalic nuclei. The lack of labeling of more rostral nuclei was not due to the death of descending neurons. Application of HRP immediately rostral to the regenerated spinal cord resulted in the labeling of a normal, and in some cases, greater than normal, number of neurons. To quantify the origin of axons in the regenerated spinal cord, electron microscopic montages of the regenerated spinal cord were made and the number of axons counted, before and after various spinal lesions. Only lesions within one spinal segment of the regenerated spinal cord had a significant effect on the number of axons in the regenerated tail spinal cord. This indicated that most of the regenerated axons were of local spinal origin. A significant increase in the number of labeled local spinal neurons was revealed following application of HRP to a regenerated tail spinal cord. These results suggest that while various portions of the lizard central nervous system can grow axons into the regenerating tail spinal cord, the great majority of axons in the regenerate are of local origin and that some of these arise from neurons that do not normally possess descending projections. Finally, to test whether new neurons were participating in the regeneration process, 3H-thymidine was injected during the regrowth of the tail. No labeled spinal cord cells were conclusively identified as neurons. Thus, the regenerating lizard tail spinal cord exhibits robust axonal sprouting from neurons near the site of a spinal transection in a manner reminiscent of sprouting in the mammalian CNS. This sprouting can develop into descending spinal projections that extend for significant distances into the regenerated tail spinal cord and provides a unique model for exploring the requirements for successful axon growth in an adult vertebrate CNS.