Rhesus monkeys were given either complete or partial dorsal rhizotomy of the cervicothoracic region and the severed stumps were left free in the dural sac. Following death 8-54 months later, serial oblique-longitudinal sections of the rhizotomied portion of the cord were examined using a silver stain that emphasized intact fibers. Afferent regeneration from the distal stumps of the severed dorsal roots into the spinal cord was found to be exceedingly rare (i.e. only one such fiber was observed). Hook-shaped fibers were observed in the proximal stumps with the locus of flexion closest to the spinal cord. These fibers suggest the presence of a barrier impeding the growth of regenerating axons. Hook-shaped fibers oriented in the opposite direction were observed within the spinal cord at its junction with the proximal stumps, again suggesting a growth barrier at the margin of the cord. Because a single oblique-longitudinal section can intercept many dorsal roots, this method provides a convenient means of verifying the completeness of rhizotomy. The results provide anatomical confirmation for electrophysiological and behavioral data indicating that following chronic dorsal rhizotomy afferent input cannot reach the spinal cord over the sectioned dorsal roots as a result of afferent regeneration. The extensive behavioral capacity exhibited by deafferented monkeys must therefore be explained on the basis of other, presumably central, mechanisms, such as central motor programs and central feedback. © 1980.