Check-All-That-Apply (CATA) is a popular method used for collecting word-based sensory descriptions from consumers. Free-Comment (FC), as a response to open-ended questions, is an interesting alternative because it removes biases due to the use of a predefined list of descriptors. In the context of a home used WA (HUT), FC enables subjects to express themselves more naturally. The present study investigated the relevance of the use of FC at home for word-based sensory description of a set of products. Two groups of 60 consumers of red wines characterised four French red wines from different terroirs performing either a CATA task or a FC task. The two sensory tasks were performed at home according to sensory modality: visual, olfactory and gustatory. The first objective was to investigate whether a FC protocol can be successfully conducted at home and whether it enables the characterisation and discrimination of a set of products. The second objective was to investigate whether extrinsic sensory information affects FC descriptions. The third objective was to investigate whether CATA and FC provide comparable information in the HUT context. The results show that an FC protocol is feasible at home and that the extrinsic sensory information did not affect FC descriptions. FC enabled better characterisation and discrimination of the products than CATA. A new WA of product differences based on the total bootstrap procedure was proposed to compare FC and CATA.