Recent pharmacological and biochemical studies have suggested that there may be more than one molecular form of the neurokinin-1 receptor (NK-1), a long and short isoform differing in the length of their cytoplasmic carboxyl-terminal tails, but no definitive evidence of the existence of such NK-1 receptor isoforms in tissue has been presented. To examine whether these different isoforms are expressed in vivo we have compared the distribution of high affinity substance P (SP) binding sites (visualized by autoradiography with [125I]SP), with the distribution of the C-terminal epitope of the full length receptor (visualized with a specific antibody against the extreme C-terminal sequence). The former method labels both long and short forms of the NK-1 receptor, while the latter labels only the long form of the protein. In the rat there is a close correspondence of [125I]SP binding and NK-1 immunoreactivity in the striatum, suggesting that the long isoform predominates in this tissue. In the parotid and submaxillary gland, there are very high levels of [125I]SP binding but only low levels of NK-1 immunoreactivity, suggesting that expression of the short form predominates in these tissues. These results imply that different tissues express different ratios of the two isoforms of the NK-1 receptor. This differential expression provides the theoretical basis for tissue specific pharmacological targeting of NK-1 receptors.