The aim of this study was to determine whether alpha(2)-adrenoceptors or imidazoline I-1-receptors are responsible for the central sympathoinhibition produced by rilmenidine and moxonidine, two clonidine-like antihypertensive drugs. Rilmenidine and moxonidine were compared with. the indirectly acting alpha(2)-adrenoceptor agonist alpha-methyldopa. Three antagonists were used. Yohimbine and SK&F86466 were used as selective alpha(2)-adrenoceptor antagonists. They were compared with efaroxan which is also an al-adrenoceptor antagonist, but, in addition, possesses affinity for imidazoline I-1-receptors. According to some but not all studies, the affinity of efaroxan for I-1-receptors is much higher than its affinity for alpha(2)-adrenoceptors. Drugs were administered into the cisterna cerebello-medullaris of conscious rabbits by a catheter implanted previously under halothane anaesthesia. Rilmenidine (10 mu g kg(-1)), moxonidine (0.3 mu g kg(-1)) and alpha-methyldopa (0.4 mg kg(-1)) lowered blood pressure and the plasma noradrenaline concentration; the degree of sympathoinhibition produced by the three agonists was very similar When inject ed after the agonists, efaroxan (0.1-14 mu g kg(-1); cumulative doses), yohimbine (0.4-14 mu g kg(-1)) and SK&F86466 (0.4-44 mu g kg(-1)) counteracted the effects of the agonists on blood pressure and the plasma noradrenaline concentration. Efaroxan was about tenfold more potent than yohimbine and SK&F86466 at antagonizing the hypotensive effects of alpha-methyldopa. Similarly, efaroxan was two- to tenfold more potent than yohimbine and SK&F86466 against rilmenidine and moxonidine, Finally, efaroxan was about as potent against alpha-methyldopa as against rilmenidine and moxonidine. The results confirm previous observations that selective alpha(2)-adrenoceptor antagonists are capable of completely antagonizing effects of rilmenidine and moxonidine. The effects of the alpha(2)-adrenoceptor antagonist with an additional high affinity for imidazoline I-1-receptors, efaroxan, can also be explained by blockade of alpha(2)-adrenoceptors. Efaroxan was more potent against rilmenidine and moxonidine than the selective alpha(2)-adrenoceptor antagonists. This was probably due to the fact that the affinity of efaroxan for alpha(2)-adrenoceptors is higher than the affinity of yohimbine and SK&F86466, since efaroxan was also the most potent of the three antagonists against the indirectly acting alpha(2)-adrenoceptor agonist a-methyldopa. The observation that efaroxan was equally potent against rilmenidine and moxonidine and against alpha-methyldopa suggests that the same receptors were involved in the effects of the three agonists, alpha(2)-adrenoceptors; this observation is not compatible with the high I-1/alpha(2) selectivity of efaroxan and the hypothesis that rilmenidine and moxonidine activate I-1-receptors, whereas alpha-methyldopa activates alpha(2)-adrenoceptors. Thus, the data do not indicate involvement of I-1 imidazoline receptors in the central sympathoinhibition elicited by rilmenidine and moxonidine in rabbits. it is likely that rilmenidine and moxonidine produce sympathoinhibition by activating the same receptors which are activated by the indirectly acting catecholamine alpha-methyldopa, namely alpha(2)-adrenoceptors.