For years now, wine writers have been tempted to describe certain wines in terms of musical metaphors and analogies. Until recently, however, it has never been altogether clear how widely shared, and hence meaningful, such surprising cross-sensory connections really were. A growing body of scientific evidence, however, now shows that regular consumers (i.e. non-experts) do reliably match certain wines with particular pieces of music (under conditions of forced choice). When questioned, people also feel that certain wines go well with specific pieces of music, while others do not. As such, it can be argued that describing wines in musical terms can potentially provide useful information concerning the likely sensory, descriptive, analytic and/or hedonic properties of the wine. While some commentators have sought for an explanation for such crossmodal matches in terms of synaesthesia, here we argue that crossmodal correspondences—the associations that the majority of us share between tastes, aromas, flavours, and mouthfeel characteristics on the one hand and particular properties of sound and music on the other—offer a more satisfactory explanation for what may be going on. In particular, we highlight how structural, statistical, semantic and affective correspondences could all play a part in explaining the affinity that so many of us feel between wine and music.