The mechanism of adenosine-induced vasodilation in rat diaphragm microcirculation was investigated using laser Doppler flowmetry. Adenosine (10⁵, 3.2 × 10⁵, and 10⁴ M), the nonselective adenosine agonist 5'-N-ethylcarboxamido-adenosine (NECA) (10⁸-10⁷ M), the specific A_2A agonist 2-p-(2-carboxyethyl)phenyl-amino-5'-N-ethyl carboxamidoadenosine (CGS-21680) (10⁸-10⁷ M), and the adenosine agonist with higher A₁-receptor affinity, R-N⁶-phenylisopropyladenosine (R-PIA) (10⁷, 3.2 × 10⁷, and 10⁶ M) elicited a similar degree of incremental increase of microcirculatory flow in a dose-dependent manner. The ATP-dependent potassium (K_ATP) channel blocker glibenclamide (3.2 × 10⁶ M) significantly attenuated the vasodilation effects of these agonists. Adenosine-induced vasodilation could be significantly attenuated by the nonselective adenosine antagonist 8-(p-sulfophenyl)-theophylline (3 × 10⁵ M) or the selective A_2A antagonist 4-(2-[7-amino-2-(2-furyl)[1,2,4]triazolo[2,3-a][1,3,5]triazin-5-ylamino]ethyl) phenol (ZM-241385, 10⁶ M), but not by the selective A₁ antagonist 8-cyclopentyl-1,3-dipropylxanthine (5 × 10⁸ M). Adenylate cyclase inhibitor N-(cis-2-phenyl-cyclopentyl) azacyclotridecan-2-imine-hydrochloride (MDL-12330A, 10⁵M) effectively suppressed the vasodilator response of adenosine and forskolin. These results suggest that adenosine-induced vasodilation in rat diaphragm microcirculation is mediated through the stimulation of A_2A receptors, which are coupled to adenylate cyclase activation and opening of the K_ATP channel.