1. Human in vitro preparations of transverse or distal colonic circular smooth muscle were potently and dose-dependently contracted by neurokinin A (EC50, 4.9 nM), the tachykinin NK2-receptor selective agonist [beta-Ala8]neurokinin A (4-10) ([beta-Ala8]NKA (4-10)) (EC50, 5.0 nM), neurokinin B (EC50, 5.3 nM) and substance P (EC50, 160 nM), but not by the tachykinin NK1-receptor selective agonist [Sar9Met(O2)11] substance P, or the NK3-receptor selective agonists, senktide and [MePhe7] neurokinin B. No regional differences between transverse and distal colon were observed in response to [beta-Ala8]NKA (4-10). 2. Atropine (1 microM) and tetrodotoxin (1 microM) did not significantly inhibit responses to [beta-Ala8]NKA (4-10), neurokinin A, substance P or neurokinin B. 3. The newly developed non-peptide antagonists for tachykinin NK2-receptors SR 48968, SR 144190 and its N-demethyl (SR 144743) and N,N-demethyl (SR 144782) metabolites, were used to challenge agonist responses, as appropriate. SR 144190 and the metabolites all potently and competitively antagonized the response to [beta-Ala8]NKA (4-10), with similar potency (Schild plot pA2 values 9.4, 9.4 and 9.3, slope = 1). SR 48968 antagonism was not competitive: the Schild plot slope was biphasic with a high (X intercept approximately 9.3) and a low (X intercept 8.4, slope 1.6) affinity site. Co-incubation of SR 48968 (10, 100 nM) and SR 144782 (10 nM) produced additive effects; in this experimental condition, SR 48968 apparent affinity (pKB) was 8.2. In addition, SR 144782 (0.1 microM) antagonized responses to neurokinin A, substance P and neurokinin B, with pKB consistent with its affinity for tachykinin NK2-receptors. The potent and selective NK1 and NK3-receptor antagonists, SR 140333 and SR 142801 (both 0.1 microM), failed to inhibit contractions induced by SP or NKB. 4. In conclusion, the in vitro mechanical responses of circular smooth muscle preparations from human colon are strongly consistent with the presence of non-neuronal tachykinin NK2-receptors, but not tachykinin NK1- or NK3-receptors. Our findings with SR 48968 suggest the existence of two tachykinin NK2-receptor subtypes, that it seems to distinguish, unlike SR 144190 and its metabolites. However, the precise nature of SR 48968 allotopic antagonism remains to be elucidated, since allosteric effects at the tachykinin NK2-receptor might well account for the complexity of the observed interaction.