The objective of the present in vitro and in vivo experiments was to examine the involvement of neurokinin NK 3 receptors in the regulation of the noradrenergic function in gerbils and guinea-pigs. Application of senktide, a peptide NK 3 receptor agonist, on guinea-pig locus coeruleus slices increased the firing rate of presumed noradrenergic neurons (EC 5 0 =26 nM) in a concentration-dependent manner. Given i.c.v., senktide (0.5-2 μg) and (MePhe 7 )neurokinin B (1-10 μg), another NK 3 receptor agonist, reduced exploratory behaviour in gerbils in a dose-dependent manner (2 μg of senktide producing a 50% reduction of locomotor activity and rearing). In vivo microdialysis experiments in urethane-anaesthetized guinea-pigs showed that senktide (2-8 μg i.c.v.) induced a dose-dependent increase in norepinephrine release in the medial prefrontal cortex. The electrophysiological, behavioural and biochemical changes elicited by senktide were concentration- or dose-dependently reduced by SR 142801, the selective non-peptide NK 3 receptor antagonist. In the locus coeruleus slice preparation, complete antagonism of senktide (30 nM) was observed with 50 nM of SR 142801, while injected i.p. (0.1-1 mg/kg) it abolished the senktide-induced norepinephrine release in guinea-pigs. In gerbils, SR 142801 (1-10 mg/kg i.p.) reversed the reduction of exploratory behaviour induced by senktide (1 μg). By contrast, the 100-fold less active enantiomer, SR 142806, did not exert any antagonism in these models. Finally, the reduction of exploratory behaviour in gerbils was found to be reversed by prazosin (0.25-256 μg/kg i.p.) and to some extent by clonidine, drugs known to depress noradrenergic function.All these experiments strongly support the hypothesis that brain noradrenergic neurons can be activated by stimulation of neurokinin NK 3 receptors.