The distribution of corticospinal projection neurons in adult rats was determined using a retrograde tracing technique. Horseradish peroxidase (HRP) and an emulsifier (Nonidet) were injected into the 5th and 6th segments of the cervical spinal cord. The greatest concentrations of HRP-positive neurons were distributed in area 4 and rostral area 6/8 (motor cortices) and medial area 3 and caudal area 2 (somatosensory cortices). The largest labeled neurons were in areas 4 and 3. HRP-positive neurons were absent or few in regions of motor and somatosensory fields which contained the face representation. Less dense concentrations of retrogradely labeled neurons were also in posterior parietal and association areas 14, 39 and 40, rostral occipital visual areas 18a and 18b, and anterior cingulate and prefrontal areas 24a, 24b, and 32. The topography of the corticospinal pathway was determined by injecting HRP without Nonidet into the cervical, upper thoracic, lower thoracic, or lumbar spinal cord. Although the distribution of labeled neurons decreased with distance down the spinal cord, the size of the corticospinal neurons in each cytoarchitectonic area was not significantly different regardless of where the injection was placed. For example, upper thoracic cord injections retrogradely labeled neurons in each of the regions containing neurons filled by cervical cord injections, however, lumbar injections retrogradely labeled neurons only in caudal areas 4 and 3 and in area 18b. The distribution of corticospinal neurons in rats in similar to the organization of the corticospinal system in higher animals. The origin of corticospinal neurons in occipital and cingulate cortices may be related to visuomotor and visceromotor control.