The use of antipsychotic drugs represents an important approach for the treatment of schizophrenia. However, their efficacy is limited to certain symptoms of this disorder, and they induce serious side effects. As a result, there is a strong demand for the development of new drugs, which depends on reliable animal models for pharmacological characterization. The present review discusses the face, construct, and predictive validity of classical animal models for studying the efficacy and side effects of compounds for the treatment of schizophrenia. These models are based on the properties of antipsychotics to impair the conditioned avoidance response and reverse certain behavioral changes induced by psychotomimetic drugs, such as stereotypies, hyperlocomotion, and deficit in prepulse inhibition of the startle response. Other tests, which are not specific to schizophrenia, may predict drug effects on negative and cognitive symptoms, such as deficits in social interaction and memory impairment. Regarding motor side effects, the catalepsy test predicts the liability of a drug to induce Parkinson-like syndrome, whereas vacuous chewing movements predict the liability to induce dyskinesia after chronic treatment. Despite certain limitations, these models may contribute to the development of more safe and efficacious antipsychotic drugs.