AbstractWe examined the pattern of cytochrome oxidase (CO), Nissl staining, and gamma‐amino butyric acid (GABA) immunoreactivity in the ventroposterior lateral nucleus (VPL) of the thalamus in monkeys that received no, total, or subtotal, ablation of the hand representations in postcentral somatosensory cortex. In unoperated animals, the region of VPL representing the hand was characterized by relatively dense and homogeneous CO staining throughout the rostral‐caudal extent of VPL. Counts of neurons in the VPL hand representation from adjacent thalamic sections processed for Nissl and GABA immunostaining indicated that there were approximately 261.4 neurons/mm2 of which 78.4/mm2 stained positive for GABA. GABA (+) puncta‐like terminals were readily apparent throughout the VPL.By contrast, animals that received total removals of the postcentral hand representations showed a dramatic reduction in CO staining in the VPL, which was confined to the expected location of the thalamic hand representation. Counts of neurons in the affected region from adjacent sections that underwent Nissl staining and GABA immunostaining also revealed a dramatic reduction of Nissl‐stained neurons, with a smaller reduction in the number of neurons staining positive for GABA. Specifically, large to medium‐sized (<180 μm2) GABA(‐) neurons were virtually eliminated in the affected portion of the VPL, and the numbers of GABA (+) neurons were significantly reduced. The remaining population of GABA(+) neurons was typically shrunken, and no GABA(+) puncta‐like terminals were observed in the affected region.The results obtained after subtotal ablation of the postcentral hand representations (only one postcentral area spared, 3b or 3a) differed from those obtained when total removals were made. Instead of virtually complete degeneration of medium‐sized to large neurons throughout the hand representation in VPL, as was the case with total removals, after partial removals, we found alternating regions in the VPL hand representation that appeared qualitatively normal, or dramatically degenerated. Thalamic sections stained with CO revealed light, moderate, and darkly stained patches of label within the hand representation in VP, depending on the type of cortical ablation. The most dramatic reduction of Nissl‐stained neurons coincided precisely with the lightest staining CO patches. Interestingly, the only statistically significant reduction in the number of GABA (+) neurons occurred in the light CO patches. In the thalamic regions coincident with the dark and moderately stained CO patches, the number of medium‐sized and large neurons decreased, but the number of GABA(+) neurons was comparable to normal. Optical density measurements of the dark patches also indicated a statistically significant difference from normal CO staining in this region.These findings indicate that following cortical damage, complex changes occur in thalamic circuitry. Interestingly, these changes are not confined to regions of the thalamus that have their major projection to the damaged region of cortex, but include thalamic regions in which the major projection is to undamaged regions of cortex. Thus, relatively restricted cortical damage has the potential to affect regions of cortex not directly damaged, through post‐injury changes in thalamic circuitry. Such changes in neuronal circuitry, especially at the thalamo cortical level, may serve as the neural substrate for the limited recovery of behavioral function often seen after strokes or head injury. © 1995 Wiley‐Liss, Inc.