Résumé : The present study aimed to ascertain the neuroanatomical basis of an influential neuropsychological model for upper limb apraxia [Rothi LJ, et al. The Neuropsychology of Action. 1997. Hove, UK: Psychology Press]. Regional cerebral blood flow was measured in healthy volunteers using H2 15O PET during performance of four tasks commonly used for testing upper limb apraxia, i.e., pantomime of familiar gestures on verbal command, imitation of familiar gestures, imitation of novel gestures, and an action-semantic task that consisted in matching objects for functional use. We also re-analysed data from a previous PET study in which we investigated the neural basis of the visual analysis of gestures. First, we found that two sets of discrete brain areas are predominantly engaged in the imitation of familiar and novel gestures, respectively. Segregated brain activation for novel gesture imitation concur with neuropsychological reports to support the hypothesis that knowledge about the organization of the human body mediates the transition from visual perception to motor execution when imitating novel gestures [Goldenberg Neuropsychologia 1995;33:63-72]. Second, conjunction analyses revealed distinctive neural bases for most of the gesture-specific cognitive processes proposed in this cognitive model of upper limb apraxia. However, a functional analysis of brain imaging data suggested that one single memory store may be used for "to-be-perceived" and "to-be-produced" gestural representations, departing from Rothi et al.'s proposal. Based on the above considerations, we suggest and discuss a revised model for upper limb apraxia that might best account for both brain imaging findings and neuropsychological dissociations reported in the apraxia literature.