Following complete transection of the thoracic spinal cord at various times during embryonic development, chick embryos and posthatched animals exhibited various degrees of anatomical and functional recovery depending upon the age of injury. Transection on embryonic day 2 (E2), when neurogenesis is still occurring and before descending or ascending fiber tracts have formed, produced no noticeable behavioral or anatomical deficits. Embryos hatched on their own and were behaviorally indistinguishable from control hatchlings. Similar results were found following transection on E5, an age when neurogenesis is complete and when ascending and descending fiber tracts have begun to project through the thoracic region. Within 48 h following injury on E5, large numbers of nerve fibers were observed growing across the site of transection. By E8, injections of horse-radish peroxidase (HRP) administered caudal to the lesion, retrogradely labelled rostral spinal and brainstem neurons. Embryos transected on E5 were able to hatch and could stand and locomote posthatching in a manner that was indistinguishable from controls. Following spinal cord transections on E10, anatomical recovery of the spinal cord at the site of injury was not quite as complete as after E5 transection. Nonetheless, anatomical continuity was restored at the site of injury, axons projected across this region, and rostral spinal and brainstem neurons could be retrogradely labelled following HRP injections administered caudal to the lesion. At least part of this anatomical recovery may be mediated by the regeneration or regrowth of lesioned axons. Although none of the embryos transected on E10 that survived to hatching were able to hatch on their own, because several sham-operated embryos were also unable to hatch, we do not attribute this deficit to the spinal transection. When E10-transected embryos were aided in escaping from the shell, they were able to support their own weight, could stand, and locomote, and were generally comparable, behaviorally, to control hatchlings. Repair of the spinal cord following transection on E15 was considerably less complete compared to embryos transected on E2, E5, or E10. However, in some cases, a degree of anatomical continuity was eventually restored and a few spinal neurons rostral to the lesion could be retrogradely labelled with HRP. By contrast, labelled brainstem neurons were never observed following E15 transection. E15 transected embryos were never able to hatch on their own, and when aided in escaping from the shell, the hatchlings were never able to stand, support their own weight or locomote.(ABSTRACT TRUNCATED AT 400 WORDS)