Animal models for complex brain disorders, such as schizophrenia, are essential for the interpretation of postmortem findings. These models allow empirical testing of hypotheses regarding the role of genetic and environmental factors, the pathophysiological mechanisms and brain circuits that are responsible for specific neural abnormalities and their associated behavioral impairment, and the effectiveness of therapeutic treatments relative to these diseases. Recently, we developed a rodent model for neural circuitry abnormalities in discrete corticolimbic subregions of subjects with major psychoses. According to our protocol, the GABA-A receptor antagonist picrotoxin is stereotaxically infused in the basolateral amygdala to mimic a GABA defect in this region that is postulated to occur in these disorders. This protocol has been tested with a number of acute and chronic time schedules. Following picrotoxin administration in the basolateral amygdala, changes in GABAergic neurons and/or terminals in hippocampal regions CA2/3 are observed, similar to those seen in major psychoses, as well as a marked reduction in GABA-receptor-mediated currents in pyramidal neurons of this region. This has established the construct and predictive validity of this model for studying limbic-lobe circuitry abnormalities. We propose that this modeling strategy may provide a valid alternative to isomorphic models of these diseases.