The objective of this investigation was to characterise the pharmacokinetic-pharmacodynamic correlation of adenosine A1 receptor partial agonists in the chronic constriction injury model of neuropathic pain. Following intravenous administration of 8-methylamino-N6-cyclopentyl-adenosine (MCPA; 10 mg/kg) and 2'deoxyribose-N6-cyclopentyl-adenosine (2'dCPA; 20 mg/kg), the time course of the effect on the mechanical paw pressure threshold was determined in conjunction with plasma concentrations. Population pharmacokinetic/pharmacodynamic analysis was applied to derive individual concentration-effect relationships. A composite model consisting of an E(max) model for the anti-hyperalgesic effect in combination with a linear model for the anti-nociceptive effect accurately described the concentration-effect relationship. For both compounds, a full anti-hyperalgesic effect was observed. The values of the EC50 for the anti-hyperalgesic effect were (mean+/-S.D.): 3170+/-1460 and 2660+/-1200 ng/ml for MCPA and 2'dCPA versus 178+/-51 ng/ml for the reference full agonist 5'deoxyribose-N6-cyclopentyl-adenosine (5'dCPA). The values of the slope for the anti-nociceptive effect were 1.9+/-0.30 and 1.2+/-0.20 g.microl/ng, respectively, versus 55+/-8 g microl/ng for 5'dCPA. Adenosine A1 receptor partial agonists behave as full agonists with regard to the anti-hyperalgesic effect in neuropathic pain, but the anti-nociceptive effect is diminished.