This article reviews recent findings concerning the feasibility, basic neurobiology, and potential functional benefits of fetal CNS tissue grafts into acute and chronic lesions of the adult spinal cord. In the rat, neuro-anatomical observations suggest that transplants into resection cavities establish neuritic projections that could functionally reunite separated rostral and caudal segments of the host spinal cord. Furthermore, some complementary electrophysiological evidence has been obtained for synaptic connectivity between host and graft neurons. In these studies, extracellular single-unit activity was evoked in fetal spinal cord (FSC) transplants by stimulating host dorsal roots that had been juxtaposed to donor tissue at the time of transplantation. In other investigations, we examined whether grafts could also establish axonal projections to appropriate areas of gray matter in the chronically injured spinal cord. For this purpose, fetal serotoninergic (5-HT) neurons were injected caudal to complete spinal cord transections that had been made 1-3 months earlier. Immunocytochemistry revealed that these cells projected their axons into gray matter regions normally innervated by bulbospinal 5-HT neurons. To investigate transplantation in a more clinically relevant lesion model, a third group of experiments involved injection of dissociated cell suspensions into acute [less than 24 h postinjury (p.i.)]), subchronic (7-10 days p.i), and chronic (greater than or equal to one month, p.i.) contusion lesions. Such grafts routinely filled areas that otherwise would have been regions of cavitation extending rostral-caudal distances of approximately 7 mm. FSC transplants in such injuries also appeared to influence some aspects of motoneuron excitability and hindlimb locomotion. More recent studies of the cat spinal cord have extended these findings in the rat by showing long-term survival (greater than 2 years) of fetal CNS allografts in recipients with either subtotal transection or compression lesions. Preliminary studies of connectivity have also shown host-graft projection patterns similar to those seen in the rat. Behavioral analyses are currently underway to examine the effects of fetal grafts in cats with chronic postcompression lesions. These observations in the rat and cat are discussed in the general context of basic biological and clinical issues relevant to the long-term objective of promoting functional improvement in the damaged spinal cord.