In this study, we investigated the biochemical mechanisms of agonist action at the G protein-coupled D2 dopamine receptor expressed in Chinese hamster ovary cells. Stimulation of guanosine 5'-O-(3-[35S]thio)triphosphate ([35S]GTPgammaS) binding by full and partial agonists was determined at different concentrations of [35S]GTPgammaS (0.1 and 10 nM) and in the presence of different concentrations of GDP. At both concentrations of [35S]GTPgammaS, increasing GDP decreased the [35S]GTPgammaS binding observed with maximally stimulating concentrations of agonist, with partial agonists exhibiting greater sensitivity to the effects of GDP than full agonists. The relative efficacy of partial agonists was greater at the lower GDP concentrations. Concentration-response experiments were performed for a range of agonists at the two [35S]GTPgammaS concentrations and with different concentrations of GDP. At 0.1 nM [35S]GTPgammaS, the potency of both full and partial agonists was dependent on the GDP concentration in the assays. At 10 nM [35S]GTPgammaS, the potency of full agonists exhibited a greater dependence on the GDP concentration, whereas the potency of partial agonists was virtually independent of GDP. We concluded that at the lower [35S]GTPgammaS concentration, the rate-determining step in G protein activation is the binding of [35S]GTPgammaS to the G protein. At the higher [35S]GTPgammaS concentration, for full agonists, [35S]GTPgammaS binding remains the slowest step, whereas for partial agonists, another (GDP-independent) step, probably ternary complex breakdown, becomes rate-determining.