Frontal cortex was removed in 1- and 30-day-old rats. When both groups reached 90 days of age, the forelimb motor/sensory cortex in the unlesioned hemisphere was injected with wheat germ agglutinin-horseradish peroxidase (WGA-HRP) or tritiated leucine. Thalamic neurons were retrogradely labeled only ipsilateral to the WGA-HRP injection site in both neonatally and juvenile-lesioned subjects. Ventrolateral (VL), ventromedial (VM), centromedial (CM), centrolateral (CL), parafascicular (PF), posteromedial (POm), and posterior (PO) thalamic nuclei were labeled. This and the demonstration of only ipsilateral thalamocortical connections at birth helped explain the marked thalamic atrophy which developed ipsilateral to neonatal frontal cortex lesions. Death of thalamic neurons after neonatal removal of their normal cortical target could be due to their failure to sprout into the opposite cortex because that cortex was already innervated by the opposite thalamus at birth. Leucine motor/sensory cortex injections in both neonatally and juvenile-lesioned subjects labeled the ipsilateral VL, VM, CM, CL, PF, POm, and PO thalamic nuclei; contralateral CM, CL, and PF thalamic nuclei; ipsilateral medial, ventral, and lateral pontine nuclei; and parts of the contralateral pontine nuclei. The ipsilateral connections were always more robust than the contralateral connections. The contralateral corticothalamic and corticopontine projections, however, were much more numerous and widespread in neonatally compared to juvenile-lesioned subjects. The greater sparing of some motor functions said to occur in neonatal compared to adult motor cortex-lesioned subjects could be due to the plasticity of corticothalamic, corticopontine, and other corticofugal pathways, but not to the plasticity of thalamocortical pathways.