1 The receptors mediating the vasodilator responses to adenosine in the isolated mesenteric arterial bed of the rat were identified by use of selective agonists and antagonists and the involvement of the endothelium was examined. 2 Adenosine-mediated dilatation of the mesentery was potentiated by the nitric oxide synthase inhibitor, N0-nitro-L-arginine methyl ester (L-NAME, 100 gM), but in contrast, removal of the endothelium substantially reduced the responses to adenosine. 3 The order of potency of adenosine receptor agonists was: 5'-N-ethylcarboxamidoadenosine (NECA) > 2-p-(-2-carboxyethyl)phenethylamino-5'-N-ethylcarboxamidoadenosine (CGS 21680) > 2-chloro-N6-cyclopentyl-adenosine (CCPA) > adenosine, suggesting the presence of A2A receptors.4 Adenosine-mediated dilatation was inhibited by the non-selective adenosine receptor antagonist, 8-phenyltheophylline (3 gM) and by the A2 receptor antagonist 8-(3-chlorostyryl)caffeine (500 nM), but was unaffected by the Al receptor antagonist, 1,3-dipropyl-8-cyclopentylxanthine (DPCPX; 10 nM).5 Reducing the pH of the perfusate to 6.8 potentiated the actions of both CGS 21680 and adenosine, but the vasodilator effects of carbachol were the same at both pH values. The adenosine response at the lower pH as at pH 7.4, was unaffected by DPCPX. The actions of the nitrovasodilator, sodium nitroprusside, were also potentiated at pH 6.8 relative to those at the higher pH value but smaller responses were obtained at the lower pH value with forskolin, a stimulator of adenylyl cyclase, than at pH 7.4. 6 It is concluded that the adenosine receptor mediating dilatation of the rat mesenteric arterial bed is of the A2A subtype, that the response, under the conditions used, is apparently partly dependent on the endothelium (but not due to the release of nitric oxide), and that the response to activation of this receptor is potentiated by a reduction in pH which is similar to that seen in ischaemic conditions.