Modelling of complex psychiatric disorders, e.g., depression and schizophrenia, in animals is a major challenge, since they are characterized by certain disturbances in functions that are absolutely unique to humans. Furthermore, we still have not identified the genetic and neurobiological mechanisms, nor do we know precisely the circuits in the brain that function abnormally in mood and psychotic disorders. Consequently, the pharmacological treatments used are mostly variations on a theme that was started more than 50 years ago. Thus, progress in novel drug development with improved therapeutic efficacy would benefit greatly from improved animal models. Here, we review the available animal models of depression and schizophrenia and focus on the way that they respond to various types of potential candidate molecules, such as novel antidepressant or antipsychotic drugs, as an index of predictive validity. We conclude that the generation of convincing and useful animal models of mental illnesses could be a bridge to success in drug discovery.