Behavioral flexibility is the ability of a subject to change its behavior according to contextual cues. In humans, Obsessive Compulsive Disorders (OCD) is characterized by repetitive behavior, performed through rigid rituals. This phenomenological observation has led to explore the idea that OCD patients may have diminished behavioral flexibility. To address this question we developed innovative translational approaches across multiple species, including human patients suffering from obsessive-compulsive disorders, and rodent genetic models of OCD to provide original data in the perspective of enlightening the neurocognitive bases of compulsive behaviors. Behavioral flexibility may be challenged in experimental tasks such as reversal learning paradigms. In these tasks, the subject has to respond to either of two different visual stimuli but only one stimulus is positively rewarded while the other is not. After this first association has been learned, reward contingency are inverted, so that the previously neutral stimulus is now rewarded, while the previously rewarded stimulus is not. Performance in reversal learning is indexed by the number of perseverative errors committed when participants maintain their response towards previously reinforced stimulus in spite of negative reward. Unsurprisingly, this behavioral task has been adapted to mice using various response modalities (T-maze, lever press, nose-poke). Using animal models of compulsive behaviors give much more possibilities to study the deficient functions and their underlying neural basis that could lead to pathological repetitive behaviors. Here we present new behavioral set-ups that we developed in parallel in human (i.e. healthy subjects and OCD patients) and mice (i.e. controls and SAPAP3-KO mice) to study the role of the behavioral flexibility as a possible endophenotype of OCD. We observed that the subjects suffering of compulsive behaviors showed perseverative maladaptive behaviors in these tasks. By comparing the results of a similar task-design in humans and mouse models we will discuss the pertinence of such translational approach to further study the neurocognitive basis of compulsive behaviors.