Dopamine D2 receptors were inactivated by N-ethoxycarbonyl-2-ethoxy-1,2,-dihydroxy-quinoline (EEDQ) (6 mg/kg i.p.). The reduction in dopamine receptors was monitored by quantitative receptor autoradiography using [125I]iodosulpiride or [3H]raclopride as radioligands. Pretreatment of male rats with haloperidol (0.3-3 mumol/kg i.p.) produced a dose-related, complete protection against the decrease in [125I]iodosulpiride binding induced by EEDQ in the dorsal and ventral striata and in all cortical areas examined. Raclopride (0.25-10 mumol/kg i.p.) produced the same pattern of effect as haloperidol but had a weaker effect. In contrast, remoxipride (1-40 mumol/kg i.p. or s.c.) only produced a partial protection against the dopamine D2 receptor inactivation by EEDQ. The results in the EEDQ test were related to the potency to block d-amphetamine-induced hyperlocomotion and the ability to induce bar-test catalepsy in the rat. The potencies in the behavioural tests were found to correspond to the in vivo occupancy for dopamine D2 receptors as evaluated by the EEDQ-induced decrease in D2 binding. However, remoxipride differed from both haloperidol and raclopride by showing a much reduced occupancy of dopamine D2 receptors at doses with behaviourally equipotent effects. The results support earlier suggestions that remoxipride in vivo may act on a subpopulation of dopamine D2 receptors.