In this study, we analyzed the antilipolytic effects of six N(6)-cyclopentyladenosine analogs in rats and developed a mechanistic pharmacokinetic-pharmacodynamic model to quantify and predict the tissue-selective action of adenosine A(1) receptor agonists in vivo. Freely moving rats received an i.v. infusion of vehicle or compound over 15 min. Arterial blood samples were taken at regular time intervals for the determination of concentrations of drugs using HPLC analysis and of nonesterified fatty acids (NEFAs). All N(6)-cyclopentyladenosine analogs that were investigated produced a significant decrease in the NEFA plasma concentration after i.v. infusion. The pharmacokinetic behavior of each ligand was described by a standard two-compartment model. The pharmacokinetic parameter estimates then were used to simultaneously fit the individual (n = 6-8) time-NEFA concentration profiles for each agonist to a physiological indirect response model in combination with the Hill equation to obtain estimates of the NEFA elimination rate constant (k(e)) and upper asymptote (fractional inhibition), midpoint location, and midpoint slope parameter (alpha, pEC(50), and n(H), respectively) of the concentration-effect relationship. Subsequently, the data were analyzed with the operational model of agonism to obtain estimates of in vivo affinity and efficacy. It was estimated that the in vivo density and/or coupling of adenosine A(1) receptors mediating antilipolytic effects is approximately 38 times higher compared with the receptors mediating bradycardia. The model predicts that it is possible to design ligands that produce significant inhibition of lipolysis and are completely devoid of cardiovascular effects in vivo.