The distributions of spinal and medullary cells projecting to the lateral cervical nucleus (LCN) have been investigated in young cats and dogs using the retrograde horseradish peroxidase (HRP) technique. Labeled spinal cells, whose axons contribute to the spinocervical tract (SCT), were found at all levels of the spinal cord ipsilateral to the injection sites. No significant differences were found between cat and dog, nor between cases with single injections at different levels of the LCN. SCT cells were found predominantly, if not exclusively, within lamina IV, with some extension into medial lamina V. No apparent mediolateral or dorsoventral density gradient was observed within lamina IV; cells of all sizes were labeled. Cells in cervical laminae I and V-VII were occasionally labeled; these, however, were considered to be propriospinal, supplying afferent fibers to the C1-2 dorsal horn. Cells of origin of spinocerebellar fibers consistently remained unlabeled in cases with restricted HRP injections and minimal fiber damage in the dorsolateral funiculus (DLF) around the injection sites. These results, therefore, corroborate and refine the findings of electrophysiological studies of the SCT and the LCN. Labeled medullary cells were located in the caudoventral and rostral portions of the dorsal column nuclei (DCN; stellate and fusiform cells), the underlying n. medullae oblongatae centralis, subnucleus dorsalis (parvicellular medullary reticular formation), the marginal and magnocellular layers (both large and small cells) of the n. trigeminalis spinalis pars caudalis and also in pars interpolaris; a cluster of cells was also consistently labeled in the lateral reticular formation just ventral to pars caudalis. The projection from the DCN to the LCN was confirmed with the anterograde Nauta technique. Fiber degeneration was observed in the entire ipsilateral LCN, although it was less abundant than that observed in the adjacent C1-2 dorsal horn. These results indicate that neurons in the rostral portions of the DCN not only may affect the input to the LCN (at the level of the dorsal horn), but also the output of the LCN itself. These data also suggest the possibility of both noxious and non-noxious facial input to the LCN.