The present study illustrates the methodology and application to evaluate the sensory data issued from an in-hall test session applied to a group of wines. It aims to provide a global vision of the sensory profiles and free descriptions of wines using the classical statistical methods and treating the two types of evaluations separately, as well as taking into account more complex structure of such data and approaching them jointly, by applying specific tools of extended multiple factor analysis methodology. Furthermore, the study focuses on precising the perceptual typicality concept from the experts point of view by highlighting the sensory attributes of wines subjected to the study percieved as typical by the panel. Results demonstrated that specific characteristics showed to be predictive of high typicality ratings for a wine, whilst others showed to be predictive of low typicality ratings.. A large panel of 43 experts has evaluated 60 wines of the DOQ Priorat through classical 24 descriptors and free comments. This project intents to capture what is behind the typicality of these wines, as expressed by the experts