Following one famous model, face and facial expression visual processing is mediated by a distributed neural system, which includes visual regions (core system) and prefrontal and limbic regions (extended system). Simulationist models suggest that, for expression recognition, along with the visual process, a nonvisual process of internal simulation of the somatovisceral and motor responses, associated with the observed emotion, is mandatory (embodied simulation). In the present review, we present these models and discuss existing evidence on the connectivity between the neural regions involved in facial expression visual processing and the simulative processes, and on how this seems to facilitate emotion recognition.