To determine whether alpha 2-adrenoceptor stimulation can augment adenosine-induced coronary vasodilation, 34 open-chest dogs were studied. When a small dose of clonidine (up to 0.24 micrograms.kg-1.min-1 ic) was administered under beta-adrenoceptor blockade, coronary blood flow [312 +/- 16 (SE) ml.100 g-1.min-1] maximally induced by intracoronary infusion of adenosine was further increased (P less than 0.05) by 66 +/- 16 ml.100 g-1.min-1, despite no significant changes in coronary perfusion pressure, myocardial oxygen consumption, and coronary venous adenosine concentration. However, when a larger dose of clonidine (0.36-0.60 micrograms.kg-1.min-1) was infused, adenosine-induced flow progressively decreased. This biphasic action of the alpha 2-adrenoceptor activity was also observed when the dose of norepinephrine was increased during alpha 1-adrenoceptor blockade with prazosin. Norepinephrine up to 0.24 micrograms.kg-1.min-1 (ic) further increased adenosine-induced coronary blood flow by 24 +/- 5% (P less than 0.001), whereas hyperemic flow was decreased by a larger dose of norepinephrine. In contrast to the alpha 2-adrenoceptor stimulation, the alpha 1-adrenoceptor stimulation (norepinephrine with yohimbine) progressively decreased coronary blood flow. Furthermore, with a small dose of clonidine, reactive hyperemic flow significantly increased compared with that without clonidine (303 +/- 13 vs. 355 +/- 13 ml.100 g-1.min-1, P less than 0.001), but a larger dose of clonidine adversely reduced reactive flow (254 +/- 18 ml.100 g-1.min-1, P less than 0.001). Adenosine release during reactive hyperemia with and without intracoronary infusions of clonidine were not altered significantly.(ABSTRACT TRUNCATED AT 250 WORDS)