To determine whether embryonic spinal cord transplants retained the ability to prevent retrograde death of Clarke's nucleus (CN) neurons if supplied after a delay, we hemisected adult rats at the T8 spinal cord segment and placed transplants of fetal tissue into the hemisection cavity immediately or up to 14 days later. Transplants provided in the first 7 days after injury prevented virtually all of the 30% loss of CN neurons at L1 ipsilateral to hemisection that occurs without a transplant. Transplants supplied at 14 days post-hemisection were ineffective. Because prevention of retrograde neuron death is one mechanism by which transplants may contribute to locomotor recovery after spinal cord injury, this window of effectiveness should be considered in the design of clinical trials.