Rats that sustain forelimb removal on postnatal day (P) 0 exhibit numerous multi-unit recording sites in the forelimb-stump representation of primary somatosensory cortex (SI) that also respond to hindlimb stimulation when cortical GABA_A+B receptors are blocked. Most of these hindlimb inputs originate in the medial SI hindlimb representation (Lane et al. 1997, 1999). Although many forelimb-stump sites in these animals respond to hindlimb stimulation, very few respond to stimulation of the face (vibrissae or lower jaw), which is represented in SI just lateral to the forelimb. The lateral to medial development of SI may influence the capacity of hindlimb (but not face) inputs to \"invade\" the forelimb-stump region in neonatal amputees. The SI forelimb-stump was mapped in adult (> 60 days) rats that had sustained amputation on either embryonic day (E) 16, on P0, or during adulthood. GABA receptors were blocked and subsequent mapping revealed increases in non-stump inputs in E16 and P0 amputees: fetal amputees exhibited forelimb-stump sites responsive to face (34%), hindlimb (10%), and both (22%); neonatal amputees exhibited 10% face, 39% hindlimb, and 5% both; adult amputees exhibited 10% face, 5% hindlimb, and 0% both, with ~80% stump-only sites. These results indicate age-dependent differences in receptive field reorganization of the forelimb-stump representation, which may reflect the spatiotemporal development of SI. Results from cobalt chloride inactivation of the SI vibrissae region and electrolesioning of the dysgranular cortex suggest that normally suppressed vibrissae inputs to the SI forelimb-stump area originate in the SI vibrissae region and synapse in the dysgranular cortex.