The antagonist affinity for a given receptor is traditionally considered to be constant, reflecting the chemical nature of the specific ligand-receptor interaction. However, recent observations with all three beta-adrenoceptors have cast doubt on this basic pharmacological principle. The extent to which this finding applies to other G protein-coupled receptors and their interaction with different G proteins is unknown. Therefore, we studied the influence of different agonists on antagonist affinity measurements for Gi- and Gs-coupled conformations of the adenosine A1-receptor in Chinese hamster ovary cells stably expressing the human adenosine A1-receptor and a cAMP-response element (CRE)-secreted placental alkaline phosphatase reporter gene. Gi-coupled inhibition of [3H]cAMP accumulation via the A1-receptor was observed at low concentrations of agonist; however, a small increase in [3H]cAMP accumulation was also seen at higher agonist concentrations. This biphasic response was more evident for A1-stimulated CRE-gene transcription. The inhibitory component was abolished by pretreatment with pertussis toxin, whereas the stimulatory component was augmented, suggesting that the responses were due to an A1-Gi-coupled inhibition followed by an A1-Gs-coupled stimulation. However, the antagonist affinity values measured at the Gi-coupled and Gs-coupled conformations of the receptor were the same in both functional responses and whole-cell binding. Thus, in marked contrast to the beta-adrenoceptors, the A1-receptor conforms to the long-held principle of pharmacology that antagonist affinity measurements are constant regardless of the response being measured and the competing agonist used to stimulate that response. This was true even when the receptor was shown, in the same assay, to exist in two different conformational states coupled to two different G proteins.