1 The aim of this study was to characterize the adenosine receptor mediating vasodilation in the microvasculature of the hamster cheek pouch in vivo. A range of adenosine agonists was used including N6-cyclopentyladenosine (CPA) (A1 agonist), 5'-N-ethylcarboxamidoadenosine (NECA) (non-selective), 2-chloroadenosine (2CADO) (non-selective), 2-p-(2-carboxyethyl)-phenethylamino-5'-N-ethylcarboxamidoadenosine (CGS 21680) (A2A agonist), N6-(3-iodobenzyl)-adenosine-5'-N-methyluronamide (IBMECA) (A3 agonist) and adenosine, as well as the adenosine antagonists 8-sulphophenyltheophylline (8-SPT) (A1/A2 antagonist), 1,3-dipropyl-8-cyclopentylxanthine (DPCPX) (A1 antagonist) and 4-(2-[7-amino-2-(2-furyl)[1,2,4]-triazolo[2,3-a][1,3,5]triazin-5-ylamino]ethyl)phenol (ZM 241385) (A2A antagonist). 2 All the adenosine analogues used induced vasodilation at concentrations between 10 nm and 1 microm, and the potency order was NECA > CGS 21680 > 2CADO > CPA=IBMECA >> adenosine, indicating an action at A2A receptors. 8-SPT (50 microm) antagonized vasodilator responses to NECA with an apparent pKB of 5.4, consistent with an action at A1 or A2 receptors and confirming that A3 receptors are not involved in this response. 3 DPCPX (10 nm) had no effect on vasodilation evoked by NECA, suggesting that this response was not mediated via A1 receptors, while ZM 241385 (10 nm) antagonized dilator responses to NECA with an apparent pKB of 8.9 consistent with an action via A2A receptors. 4 Overall these results suggest that adenosine A2A receptors mediate vasodilation in the hamster cheek pouch in vivo.