The present study used plethysmometry to examine oedema following local injection of selective adenosine A(1), A(2) and A(3) receptor agonists and inhibitors of adenosine metabolism into the hindpaw of the rat. N(6)-Cyclopentyladenosine and L-N(6)-phenylisopropyladenosine (A(1)), 2-[p(2-carboxyethyl) phenethylamino]-5'-N-ethylcarboxamidoadenosine hydrochloride (CGS21680) (A(2A)) and N(6)-benzyl-5'-N-ethylcarboxamido adenosine (N(6)-B-NECA) (A(3)) all produced an increase in paw volume (N(6)N(6)-cyclopentyladenosine, L-N(6)CGS21680). At the highest dose, each agent also produced a systemically mediated suppression of oedema. Oedema by N(6)-cyclopentyladenosine was blocked by caffeine, 8-cyclopentyl-1,3-dimethylxanthine and enprofylline. Oedema by CGS21680 was blocked by caffeine and 8-cyclopentyl-1, 3-dimethylxanthine. Oedema by N(6)-B-NECA was blocked by enprofylline, but not by caffeine or 8-cyclopentyl-1, 3-dimethylxanthine, or by systemic administration of MRS 1191. Oedema by both N(6)-cyclopentyladenosine and N(6)-B-NECA was blocked by mepyramine, ketanserin and phentolamine, but that by CGS21680 was not. The adenosine kinase inhibitor 5'-amino-5'-deoxyadenosine and the adenosine deaminase inhibitor 2'-deoxycoformycin produced only a limited increase in paw volume, and this was blocked by caffeine. This study demonstrates an acute paw oedema response following local administration of adenosine A(1), A(2) and A(3) receptor agonists, which likely results from different mechanisms of action in each case.