Using immunofluorescence and cytofluorimetric scanning techniques in the rat, the fast anterograde and retrograde axonal transport of growth-associated protein-43-like immunoreactivity in normal sciatic nerves, and after spinal cord transection in the lower thoracic region, were investigated. Spinal roots and motor endplates in the peroneal muscles were also studied. For comparison, anti-synaptophysin (p38) was used. In intact adult animals, the amounts of immunoreactive growth-associated protein-43 increased linearly, both proximally and distally to the crush site, between 1 and 24 h after crushing the sciatic nerve. The accumulations were present in thick as well as in thin axons. Distal accumulations in the sciatic nerve were about 40-60% of the proximal amounts, indicating a recycling of organelles with growth-associated protein-43-like immunoreactivity. During the week after spinal cord transection, no clear changes were observed; the anterograde transport of growth-associated protein-43-like immunoreactivity showed a tendency to decrease at day 1 and then a tendency to increase, reaching 120% of control at seven days (not significant). Transported p38-like immunoreactivity showed similar but smaller changes. In the lumbar spinal cord gray matter many nerve terminals with growth-associated protein-43-like immunoreactivity were seen in intact animals. After spinal transection, these terminals gradually decreased, suggesting that they belonged to descending pathways. However, p38-positive terminals were not obviously decreased. After crushing ventral and dorsal roots, accumulations of pf growth-associated protein-43-like immunoreactivity were present in thick axons in the ventral roots and in thin to medium-sized axons in the dorsal roots. In peroneal muscles, growth-associated protein-43-like immunoreactivity was present in some (but not all) motor endplates in all groups. These results indicate that: (i) growth-associated protein-43 is normally present in nerve terminals of many descending projections of the spinal cord; (ii) growth-associated protein-43-like immunoreactivity is expressed and bidirectionally transported in neurons (motor as well as sensory) of normal sciatic nerves; (iii) growth-associated protein-43-like immunoreactivity is present in some adult motor endplates; and (iv) inhibited supraspinal input causes minor, if any, alterations--paralleled by p38--in axonal transport of growth-associated protein-43-like immunoreactivity.