When applied peripherally, adenosine has been shown to be pronociceptive in a number of animal and human models. Recent evidence has implicated adenosine as a significant mediator in the inflammatory process. In this study using rats, we have examined the effect of adenosine and of selective adenosine A(1) and A(2) receptor agonists and antagonists on the response to a subcutaneous injection of formalin into the rat hindpaw. Adenosine co-injected with formalin 0.5% significantly increased flinching in both phases in a dose-dependent manner. The highest dose of adenosine had no behavioral effect on its own. The adenosine A(2) receptor agonist 2-p-(2-carboxyethyl)phenethylamino-5'-N-ethylcarboxamido adenosine hydrochloride (CGS-21680), at a dose of 1.5 nmol, increased flinching associated with 0.5% formalin injection but at higher doses produced depressant effects due to systemic absorption. The adenosine A(1) receptor agonist N-6-cyclohexyladenosine produced only systemic behavioral effects as determined by contralateral application. The flinching response to 2.5% formalin was significantly decreased by the adenosine A(2) receptor antagonist 3,7-dimethyl-1-propargylxanthine (DMPX). In contrast, 8-cyclopentyl-1,3-dimethylxanthine (CPT), the selective adenosine A(1) receptor antagonist augmented the response to 2.5% formalin. The non-selective adenosine receptor antagonist caffeine had no significant effect over a wide range of doses. In summary, exogenous adenosine enhances nociception in the formalin test, probably via a peripheral A(2) receptor-mediated action. Endogenous adenosine, acting at both A(1) and A(2) receptors, appears to be involved in the formalin-induced inflammatory response. This activation of adenosine A(1) and A(2) receptors may have opposing effects on nociceptive input.