The selective non-peptide NK(1) receptor antagonist NKP608 has been shown to exert potent anxiolytic-like effects in the rat and gerbil social interaction tests. In vitro binding of NKP608 in cortical, striatal and rest-of-brain tissue samples from mice, rats and gerbils indicated comparable pIC(50) values for rats and mice (in all three tissues) and only slightly higher values for gerbils. It would therefore be expected that doses previously found to produce anxiolytic-like effects in rats and gerbils would also be active in mice. The present study evaluated NKP608 in one of the most widely-used animal models of anxiety, the mouse elevated plus-maze. Two consecutive experiments were conducted in which the effects of NKP608 (0.0003-10.0 mg/kg, p.o.) were compared to those produced by the prototypical benzodiazepine anxiolytic, chlordiazepoxide (CDP, 15 mg/kg, p.o.). Ethological scoring methods were used to provide comprehensive behavioural profiles for each compound. In both experiments, acute CDP treatment resulted in significant anxioselective effects, i.e., reductions in measures of open arm avoidance without any alteration in general activity levels (closed arm entries and rearing). Although the results of Experiment 1 (0.001-10.0 mg/kg NKP608) suggested a weak anxiolytic-like action of NKP608 at 0.001 mg/kg (significant increase in percent open arm entries), Experiment 2 failed both to replicate this effect or to find any behavioural activity at lower (0.0003 mg/kg) or higher (0.03 mg/kg) doses. Present findings suggest that the anxiolytic efficacy of this NK(1) receptor antagonist may be test-specific and thus limited to particular subtypes of anxiety. These new data are also discussed in relation to the general difficulty of relating the behavioural profiles of NK(1) receptor antagonists to their potency at NK(1) receptors.