Developing animal models for psychiatric disorders is a complicated process. This is principally due to our relatively limited knowledge of the processes underlying such complicated illnesses as depression and schizophrenia. Unfortunately this can lead to a vicious circle. Our limited knowledge of the disease leads to the inability to develop proper animal models, which will lead to an inability to increase our knowledge about neuronal mechanisms underlying the illness. In the present review we have tried to show that with respect to schizophrenia the situation is slowly changing. Using electrophysiological and psychological methods, clinicians begin to understand the psychopathological processes underlying the schizophrenic process. Preclinical researchers have tried to use this knowledge to develop animal models in which hypotheses can be tested and possible insight into mechanisms underlying the disease can be gained. In the present review three different animal models are presented. These are based on the construct that schizophrenic patients are disturbed in their information processing. More precisely the first two animal models are based on the construct that schizophrenic patients are less able to differentiate between relevant and irrelevant stimuli. The analysis of the literature suggests that the amphetamine-induced distruption of latent inhibition (and probably blocking), and the phencyclidine-induced disturbances in the startle response might provide two interesting animal models with construct, face and predictive validity for schizophrenia. This third model deals with the amphetamine-induced changes in the behaviour of socially living monkeys. This model seems to be related to both positive and negative symptoms of schizophrenia and is based on the construct that the negative symptoms are due to a compensatory mechanism which protects the subjects from "sensory flooding". The amphetamine-induced changes in behaviour of socially living monkeys seem to represent an example of an animal model in which both positive symptoms (stereotypy) and negative symptoms (social isolation) occurs. The construct validity of this model is still unclear, but pharmacological studies suggest that, apart from the face validity, the model also has a certain predictive validity. The models discussed in the present review can help us to increase our insight in neuronal structures underlying information processing disturbances. Structures known to be implicated in these models include the hippocampus, amygdala and ventral striatum.