Recently, a standardized program for training of affect recognition (TAR) was developed which has demonstrated efficacy and specificity with respect to behavioral performance. The effects of the TAR on the cerebral correlates were evaluated using repeated fMRI event-related measurements in a group of schizophrenia patients (n=10) before and after TAR treatment six weeks apart. A second patient group without training (n=10, treatment as usual, TAU) as well as healthy subjects (n=10) were investigated at equivalent time points. Schizophrenia patients were shown to be differentially impaired in the identification of the emotional aspects of facial expressions (but not age discrimination) when compared with healthy participants. A specific improvement in the increased number of correct identifications was observed in trained patients only. In parallel, an increase in activation was noted in the left middle and superior occipital lobe, the right inferior and superior parietal cortex, and the inferior frontal cortex bilaterally in TAR patients compared to the TAU group. These activation changes in TAR patients correlated with their behavioral improvement, further corroborating the positive effect of training. Specific training effects are seen to correspond with cerebral effects, probably reflecting a more efficient use of attentional, perceptual, or cognitive strategies.