Schizophrenia is a devastating disorder with a broad range of positive (thought disorder, delusions, and hallucinations), negative (anhedonia, poverty of speech, and social withdrawal), and cognitive symptoms. However, many of these symptom domains remain poorly treated despite the fact that the first antipsychotic drug (APD) was developed over 50 years ago. One of the difficulties of discovering new and more efficacious APDs is reproducing the signs and symptoms of this complex, very human disease in animals. This chapter provides an overview of the animal models that have been, and continue to be, used in screening putative APDs. These have typically exploited the behavioral modifications (mainly hyperactivity) induced by dopaminergic agonists or N-methyl-d-aspartate (NMDA)/glutamatergic antagonists. As such, these models have mainly addressed the florid (positive) aspects of schizophrenia. We also discuss more innovative models that have appeared over the last 10–15 years, which attempt to better reproduce the complexity of the clinical disorder, in particular by targeting negative symptoms and cognitive deficits, as well as manipulating neurotransmitter systems other than the dopaminergic or glutamatergic ones. In addition, several models have exploited human epidemiological information, and have incorporated, for example, prenatal infection that confers an increased risk of developing schizophrenia. As schizophrenia is known to have an important genetic component, there have also been attempts to create transgenic animals carrying identified risk mutations for schizophrenia. To date these models have been little used for testing putative antipsychotics. However, it seems likely that we will need to combine several of these approaches (e.g., a neurodevelopmental insult on a specific genetic background followed by assessment in a complex behavioral task related to specific symptom domains) in order to identify truly novel therapeutic agents for this disorder.