Astringency is an everyday sensory experience best described as a dry mouthfeel typically elicited by phenol-rich alimentary products like tea and wine. The neural correlates and cellular mechanisms of astringency perception are still not well understood. We explored taste and astringency perception in human subjects to study the contribution of the taste as well as of the trigeminal sensory system to astringency perception. Subjects with either a lesion or lidocaine anesthesia of the Chorda tympani taste nerve showed no impairment of astringency perception. Only anesthesia of both the lingual taste and trigeminal innervation by inferior alveolar nerve block led to a loss of astringency perception. In an in vitro model of trigeminal ganglion neurons of mice, we studied the cellular mechanisms of astringency perception. Primary mouse trigeminal ganglion neurons showed robust responses to 8 out of 19 monomeric phenolic astringent compounds and 8 polymeric red wine polyphenols in Ca(2+) imaging experiments. The activating substances shared one or several galloyl moieties, whereas substances lacking the moiety did not or only weakly stimulate responses. The responses depended on Ca(2+) influx and voltage-gated Ca(2+) channels, but not on transient receptor potential channels. Responses to the phenolic compound epigallocatechin gallate as well as to a polymeric red wine polyphenol were inhibited by the Gαs inactivator suramin, the adenylate cyclase inhibitor SQ, and the cyclic nucleotide-gated channel inhibitor l-cis-diltiazem and displayed sensitivity to blockers of Ca(2+)-activated Cl(-) channels.