To characterize adenosine-mediated vascular responses, synthetic A1 and A2 receptor agonists (N-ethyl carboxamido adenosine [NECA], 2-chloro adenosine [2CA], or cyclohexyl adenosine [CHA]), the parent compound (adenosine [ADO]), an uptake inhibitor (dipyridamole [DIPYRID]) or a nonselective, competitive antagonist (8-phenyl theophylline [8pTHEO]) were topically applied to 20-60 microns arterioles in the subcutaneous microcirculation of the hamster. Blood flow was calculated from arteriolar diameter and red blood cell velocity using intravital microscopy. At greater than 10(-8) M, the potency order for vasodilation (maximum, 170-190% of control) was NECA greater than 2CA greater than ADO; these responses were attenuated by 10(-5) M 8pTHEO. From 10(-8) to 10(-6) M, 2CA evoked vasodilation whereas ADO, which has an identical affinity at A1 and A2 receptors, evoked lesser responses. ADO-induced vasodilation was potentiated by 10(-5) M DIPYRID; this response was similar to that evoked by 2CA alone or 2CA + DIPYRID. In contrast to ADO, 2CA is a poor substrate for cellular uptake, which suggests that uptake reduces the A2 effect of exogenous ADO. From 10(-10) to 10(-8) M, CHA and ADO were equipotent antagonized by 8pTHEO. Norepinephrine was a more potent vasoconstrictor and 8pTHEO did not alter these responses. Since ADO is a metabolic substrate and a nonselective receptor agonist, while CHA is A1-selective and a poor substrate for cellular uptake, neither A2 activation nor cellular uptake altered expression of the A1 effect of exogenous ADO. Furthermore, DIPYRID had no effect on the A1 response.(ABSTRACT TRUNCATED AT 250 WORDS)