The dopamine (DA) D-1 agonist SK&F 38393 as well as the D-2 agonist pergolide and the mixed D-1/D-2 agonist apomorphine induced strong hypermotility and oral stereotypy in rats pretreated with a daily dose of reserpine for 2 and in particular for 4 days (3 and 5 injections, respectively). SK&F 38393 had no behavioural stimulant effect in saline-pretreated rats, whereas pergolide and apomorphine produced stimulation, although only after higher doses. Agonists at 5-HT and muscarinic receptors and at alpha 1-adrenoceptors were ineffective in reserpine-pretreated rats whereas the alpha 2-adrenoceptor agonist, clonidine, and the muscarinic antagonist, scopolamine, produced weak locomotor stimulation. The hypermotility induced by SK&F 38393 in reserpinized rats was blocked by pretreatment with the DA D-1 antagonists, SCH 23390 and SK&F 83566c, whereas the DA D-2 antagonists, YM 09151-2, clebopride and spiroperidol were weak or ineffective. In contrast pergolide-induced hypermotility was blocked by low doses of the D-2 antagonists but was weakly or not influenced by the D-1 antagonists. Selectivity ratios between drug potencies in the two models ranged from 65 to more than 600. The mixed D-1/D-2 antagonists, cis(Z)-flupentixol and cis(Z)-clopenthixol, blocked the effect of both SK&F 38393 and pergolide. The alpha 1-adrenoceptor antagonist, prazosin, and the 5-HT2 receptor antagonist, ketanserin, did not modify the effect of SK&F 38393 or pergolide. Stereotyped behaviour induced by a high pergolide dose in normal rats was, in contrast to the effect in reserpinized rats, blocked by low doses of either SCH 23390 or spiroperidol. Finally, the hypermotility induced by apomorphine in reserpinized rats was markedly antagonized by both SCH 23390 and spiroperidol. The results suggest a close relation between D-1 and D-2 receptor sites in normal rats. After prolonged reserpine treatment, the D-1 agonist acquires full DA agonist efficacy. Furthermore, behavioural stimulation under these conditions is mediated by two separate D-1 and D-2 receptor sites which can be manipulated independently by antagonists. The mechanism by which this phenomenon occurs is unknown but the adaptational changes show close similarities to those observed after 6-hydroxyDA-induced denervation.