A short period of vibrissae deprivation in an adolescent (≃ 1 month old) rat can lead to depression of the cortical response to stimulation of the regrown vibrissae. In a barrel column representing the deprived vibrissa, depression is greater for neurons located close to the barrel column representing the spared vibrissa. One possible explanation is that the spared vibrissa produces heterosynaptic depression of the principal vibrissa response. To test this idea further, we compared the effect of depriving all vibrissae (no heterosynaptic influence at all) with depriving a single vibrissa (maximal heterosynaptic influence expected). In addition we tested the origin of the depression by recording from subcortical structures. After 7 days' deprivation and 6-8 days' regrowth, we tested the responses of barrel cortex cells, thalamic VPm neurons and trigeminal ganglion cells to stimulation of the regrown vibrissae. We found that depression was greater in cortex if a single vibrissa had been deprived than if all vibrissae had been deprived. (Average principal vibrissae responses in single vibrissae deprived animals were 36% of those in all vibrissae deprived animals for layer II/III and 41% for layer IV.) This implicates the spared vibrissae in actively down-regulating responses to the deprived vibrissae. However, some depression could also be produced in animals deprived of all vibrissae (layers II/III were 39% and layer IV 74% of control levels). These results indicate that simple withdrawal of activation has a depressive effect on responses but that depression is far greater if some active inputs remain. Neither form of deprivation had an effect on responses to principal vibrissa stimulation in the thalamus or trigeminal ganglion however, suggesting that depression originates in the cortex. Within the cortex, intracortical connections seem most affected as the greatest depression was found in layers II/III and in layer IV among cells responding at intermediate latencies (9-14 ms).