1. Sodium ions inhibit spontaneous G(i)/G(o)-coupled receptor activity and promote agonist-induced responses in vitro. The effects of sodium on the relative efficacy of opioid agonists for G-protein activation was measured by guanosine-5'-O-(gamma-(35)S)-triphosphate ([(35)S]-GTPgammaS) binding in membranes from two mu-opioid receptor-containing systems: CHO cells stably transfected with mouse mureceptors (mMOR-CHO cells) and rat thalamus. 2. NaCl inhibited basal [(35)S]-GTPgammaS binding in both systems, and this effect was partially mimicked by KCl. In mMOR-CHO membranes, net [(35)S]-GTPgammaS binding stimulated by partial but not full agonists was inhibited by NaCl with a potency that was inversely proportional to agonist efficacy. Monovalent cations were required for agonist-stimulated [(35)S]-GTPgammaS binding in this system, and increasing NaCl concentrations magnified relative efficacy differences among agonists. 3. In thalamic membranes, which contain a lower receptor:G-protein ratio than mMOR-CHO cells, similar monovalent cation effects were observed, with two exceptions: (1) [(35)S]-GTPgammaS binding stimulated by both full and partial agonists was inhibited by NaCl; and (2) monovalent cations were not required to observe agonist-stimulated [(35)S]-GTPgammaS binding. 4. Basal [(35)S]-GTPgammaS binding stimulated by the absence of monovalent cations resembled that of agonist-stimulated binding and was blocked by pretreatment of mMOR-CHO cells with pertussis toxin. 5. These results indicate that sodium inhibits spontaneous and agonist-occupied mu receptor-mediated G-protein activation in a manner inversely proportional to the efficacy of the agonist, and that spontaneous mu receptor activity and the relative efficacy of partial agonists acting at these receptors are both increased by increases in the stoichiometric ratio of receptors:G-proteins.