1. Inhibition of NK3 receptor agonist-induced contraction in the rabbit isolated iris sphincter muscle was used to assess the in vitro functional activity of three 2-phenyl-4-quinolinecarboxamides, members of a novel class of potent and selective non-peptide NK3 receptor antagonists. In addition, an in vivo correlate of this in vitro response, namely NK3 receptor agonist-induced miosis in conscious rabbits, was characterized with some of these antagonists. 2. In vitro senktide (succinyl-[Asp9,MePhe8]-substance P (6-11) and [MePhe7]-neurokinin B ([MePhe7]-NKB) were potent contractile agents in the rabbit iris sphincter muscle but exhibited quite different profiles. Senktide produced monophasic log concentration-effect curves with a mean pD2=9.03+/-0.06 and mean nH=1.2+/-0.02 (n=14). In contrast, [MePhe7]-NKB produced shallow log concentration-effect curves which often appeared biphasic (nH=0.54+/-0.04, n=8), preventing the accurate determination of pD2 values. 3. The contractile responses to the NK3 receptor agonist senktide were antagonized in a surmountable and concentration-dependent manner by SB 223412 ((-)-(S)-N-(alpha-ethylbenzyl)-3-hydroxy-2-phenylquinoline-4-ca rboxamide; 3-30 nM, pA2=8.4, slope=1.8+/-0.3, n=4). SB 222200 ((-)-(S)-N-(alpha-ethylbenzyl)-3-methyl-2-phenylquinoline-4-car box amide; 30-300 nM, pA2=7.9, slope=1.4+/-0.06, n=4) and SB 218795 ((-)-(R)-N-(alpha-methoxycarbonylbenzyl)-2-phenylquinoline-4-carboxamide; 0.3 and 3 microM apparent pKB=7.4+/-0.06, n=6). 4. Contractile responses to the NK3 receptor agonist [MePhe7]-NKB in the rabbit iris sphincter muscle were unaffected by SB 218795 (0.3 and 3 microM, n=8). In contrast, SB 223412 (30 and 300 microM n=4) and SB 222200 (0.3 and 3 microM, n=4) inhibited responses to low concentrations (< or = 1 nM), to a greater extent than higher concentrations (> 1 nM) of [MePhe7]-NKB. Furthermore, log concentration-effect curves to [MePhe7]-NKB became steeper and monophasic in the presence of each antagonist. 5. SB 218795 (3 microM, n=4) had no effect on contractions induced by transmural nerve stimulation (2 Hz) or substance P, exemplifying the selectivity of this class of antagonist for functional NK3 receptors over NK1 receptors in the rabbit. 6. In vivo, senktide (1, 10 and 25 microg i.v., i.e. 1.2, 11.9 and 29.7 nmol, respectively) induced concentration-dependent bilateral miosis in conscious rabbits (maximum pupillary constriction=4.25+/-0.25 mm; basal pupillary diameter 7.75+/-0.48 mm; n=4). The onset of miosis was within 2-5 min of application of senktide and responses lasted up to 30 min. Responses to two i.v. administrations of 25 microg senktide given 30 min apart revealed no evidence of tachyphylaxis. Topical administration of atropine (1%) to the eye enhanced pupillary responses to 25 microg senktide. This was probably due to the mydriatic effect of atropine since it significantly increased baseline pupillary diameter from 7.0+/-0.4 mm to 9.0+/-0.7 mm (n=4), thereby increasing the maximum capacity for miosis. Senktide-induced miosis was inhibited by SB 222200 (1 and 2 mg kg[-1], i.v., i.e. 2.63 and 5.26 micromol kg[-1]; maximum inhibition 100%; n=3-4), SB 223412 (0.5 and 1 mg kg[-1], i.v., i.e. 1.31 and 2.61 micromol kg[-1]; maximum inhibition 100%; n=3), SB 218795 (0.5 and 1 mg kg[-1] i.v., i.e. 1.26 and 2.52 micromol kg-1; maximum inhibition 78%; n=3), and the structurally distinct NK3 receptor antagonist SR 142801 ((S)-(N)-(1-(3-(1-benzoyl-3-(3,4-dichlorophenyl)piperidin-3-yl)propyl)-4-phenylepipiperidin-4-yl)-N-methylacetamide; 1.5mg kg-1, i.v., i.e. 2.47micromol kg-1, maximum inhibition 92%; n=3). 7. Topical administration of senktide (25microg; 29.7nmol) to the eye induced unilateral miosis in the treated eye only. At this dose there was no significant difference (P<0.05) between pupillary constriction obtained by topical or i.v. senktide, and topically administered atropine had no significant effect on responses to topical senktide (n=4). 8. [MePhe7]-NKB (125, 250 and 500microg, i.v., i.e. 98.31, 196.62 and 393.24nmol, respectively) also induced bilateral miosis in conscious rabbits (maximum pupillary constriction=4.13+/-0.30mm; n=4), but in contrast to in vitro studies this agonist was approximately 100 fold less potent than senktide. [MePhe7]-NKB-induced miosis was inhibited by SB 222200 (5mg kg-1, i.v., i.e. 13.14micromol kg-1; maximum inhibition 69%; n=3). 9. In summary, SB 223412, SB 222200 and SB 218795 are potent and selective antagonists of NK3 receptor-mediated contraction in the rabbit isolated iris sphincter muscle. In addition, NK3 receptor agonist-induced miosis in conscious rabbits is a good in vivo correlate of the in vitro rabbit iris sphincter muscle preparation and appears to be a useful model for characterizing the pharmacodynamic profile and efficacy of structurally distinct NK3 receptor antagonists, such as SB 222200, SB 223412, SB 218795 and SR 142801.