The dopamine agonist 6,7-ADTN induced locomotor stimulation after injection into the nucleus accumbens of rats. This hyperactivity was stereospecifically antagonized by peripheral injection of typical and atypical neuroleptics, except sulpiride, regardless of structural class. The potency of 14 neuroleptics correlated significantly to that in other neuroleptic in vivo models in rats, i.e. cataleptogenic and amphetamine-antagonistic (stereotypy, circling) activity. Sulpiride was inactive after peripheral injection, whereas intraacumbens sulpiride antagonized 6,7-ADTN-induced hyperactivity with a higher potency than that found for locally injected cis-(Z)-flupentixol, haloperidol and clebopride. Blockade of alpha 1-adrenergic receptors by systemic prazosin or locally injected phentolamine also antagonized 6,7-ADTN, whereas antagonists of alpha 2-adrenergic, beta-adrenergic, muscarinic and serotoninergic receptors were without effect. Furthermore, the cis-(Z)-flupentixol and haloperidol-induced 6,7-ADTN blockade was only slightly or not reversed by additional antimuscarinic (scopolamine) or antiserotonergic (methysergide) treatment. It is concluded that antagonism of 6,7-ADTN-induced hyperactivity is a relevant model for detecting neuroleptic activity with minimal influence of anticholinergic and antiserotonergic activity, but that the dopamine receptor within the nucleus accumbens responsible for the locomotor stimulation is similar to that involved in other behavioural effects of neuroleptics. It is suggested that an apparent preferential limbic action of some neuroleptics (e.g. clozapine and chlorprothixene) is more likely to be explainable by their simultaneous blocking effect on dopaminergic, alpha-adrenergic and muscarinic receptors.