The receptor-coupling efficiency for epinephrine (EPI) stimulated heterotrimeric G protein activation was studied at the G protein level in membranes prepared from PC12 cells expressing cloned alpha 2D-adrenergic receptors (alpha 2D-AR). After pretreatment with different concentrations of N-ethoxycarbonyl-1,2-dihydroquinoline, which irreversibly inactivates alpha 2D-AR, the portion of alpha 2D-ARs remaining active (q) was estimated from EPI-stimulated [35S]GTP gamma S binding. This function-derived estimate was close to the actual remaining number of receptors, as determined in saturation-binding studies using the selective alpha 2-AR antagonist [3H]rauwolscine in the same membranes. The agonist dissociation constant (KA) derived from EPI-stimulated [35S]GTP gamma S binding via Furchgott analysis was similar to the EC50 of EPI in the same assay, but 40-fold lower than its Ki measured from EPI competition for [3H]rauwolscine-binding sites in the presence of GTP gamma S and Na+. The occupancy-response relationship, calculated using Ki rather than KA, was markedly nonlinear, consistent with the high expression of alpha 2D-AR in these membranes. A nonlinear occupancy-response relationship was more directly confirmed by measuring the maximal level (i.e., full occupancy level) of G protein activation at graded densities of alpha AD-AR after N-ethoxycarbonyl-1,2-dihydroquinoline treatment. Determination of the number of G-proteins activated per receptor yielded lower values at higher receptor densities, indicating that overexpression of receptors can reduce their efficiency. Our results indicate the potential utility of using GTP-binding studies to assess agonist efficacy at the G protein level under conditions where receptor occupation can also be directly measured.