This study examined the ability of an adenosine kinase inhibitor (5'-amino-5'-deoxyadenosine; NH2dAD), an adenosine deaminase inhibitor (2'-deoxycoformycin), and combinations of these agents to produce a peripheral modulation of the pain signal in the low concentration formalin model. Drugs were administered in combination with 0.5% formalin, or into the contralateral hindpaw to test for systemic effects, and episodes of flinching behaviors determined. Coadministration of NH2dAD 0.1-100 nmol with formalin produced antinociception as revealed by an inhibition of flinching behaviors. This action was peripherally mediated as it was not seen following contralateral administration of the NH2dAD, and was due to accumulation of adenosine and activation of cell surface adenosine receptors as it was blocked by the adenosine receptor antagonist caffeine. Antinociception was intensity-dependent, as it was not seen when higher concentrations of formalin (0.75%, 1.5%) were used. The coadministration of the selective adenosine A1 receptor antagonist 8-cyclopentyl-1,3-dimethylxanthine revealed the presence of an inhibitory tone of adenosine when the intrinsic antinociceptive effect of NH2dAD was obscured by the solvent or the stimulus intensity. 2'-Deoxycoformycin 0.1-100 nmol did not produce any intrinsic effect, but 100 nmol coadministered with low concentrations of NH2dAD, which lacked an intrinsic effect, augmented antinociception by NH2dAD. Again, this was a peripheral rather than a systemic response. The combined action of the adenosine kinase and deaminase inhibitors was completely reversed by coadministration of caffeine. Antinociception with NH2dAD is observed at higher concentrations of formalin in second trial experiments. This study demonstrates a peripheral antinociceptive action mediated by endogenous adenosine which accumulates following the peripheral inhibition of adenosine kinase; this action is due to activation of an adenosine A1 receptor.