BACKGROUND Obsessive compulsive disorder (OCD) is a debilitating condition with limited treatment options. OCD is heterogeneous with respect to the content of obsessions and compulsions and their underlying motivation, among other characteristics. Animal models have provided important insights into the pathophysiology of OCD. METHODS The phenomenology of OCD is discussed, with emphasis on clinically-relevant subgroups. The paper also discusses the advantages and limitations of animals as models of OCD, along with considerations on assessing their validity. A PubMed database search using the terms 'animal model' and 'obsessive compulsive disorder' revealed ongoing studies in several models, including stereotypy in the deer mouse, quinpirole-induced checking, spontaneous alternation, compulsive lever pressing, genetic models, pathogenic models and models involving normal compulsive-like behavioral patterns. These models are presented with respect to their similarity to specific features of OCD and the information gained from them. Studies in many of these models point to the participation of corticostriatal thalamocortical circuitry and corticostriatal glutamate neurotransmission in the pathophysiology of compulsive-like behavior. CONCLUSIONS The use of animal models takes us beyond simple serotonin- or dopamine-based models of OCD that are founded on the often limited, and still unexplained, response of OCD symptoms to serotonin reuptake inhibitors or antipsychotic therapy. Pharmacological challenges that selectively target neurochemical systems that modulate either corticostriatal glutamate or striatal dopamine neurotransmission, or indeed both, should be investigated in animal models of compulsive-like behavior. Such systems include metabotropic glutamate, adenosine and endocannabinoid receptors, among others.