The abilities of 4 dopamine agonists to inhibit the tonic single unit activity of substantia nigra dopamine neurons and stimulate tonic activity of globus pallidus neurons were compared to study the agonists' effects on pre- and postsynaptic dopamine receptors, respectively. The agonists studied were apomorphine and pergolide, which interact with both D1 and D2 receptors, and the selective D2 agonists quinpirole and RU 24926. Drugs were administered systemically. The 4 dopamine agonists were equipotent and equiefficacious at inhibiting the firing rates of dopamine neurons. In contrast, their effects on pallidal cells were not identical; apomorphine and pergolide induced significantly greater increases in pallidal cell activity than did quinpirole and RU 24926. In addition, pretreatment with a small dose of quinpirole did not attenuate the excitatory effect of apomorphine on globus pallidus cell activity, as low doses of apomorphine have previously been shown to do. Possible mechanisms underlying the differences in efficacy between the non-selective and D2 selective dopamine agonists in the globus pallidus were investigated. Coadministering quinpirole with apomorphine did not significantly attenuate the effect of apomorphine, suggesting that quinpirole is not a partial agonist at postsynaptic dopamine receptors. In addition, prazosin pretreatment did not attenuate the stimulatory effect of pergolide on firing rates of pallidal cells, indicating that the greater efficacy of the non-selective agonists was not due to concurrent stimulation of alpha 1 adrenergic receptors and dopamine receptors. However, the effect of quinpirole on pallidal cell activity was significantly potentiated by pretreatment with the D1 agonist RS-SKF 38393 but not its inactive enantiomer S-SKF 38393. These results suggest that concurrent D1 and D2 receptor stimulation may be necessary for the full expression of postsynaptic receptor-mediated effects of dopamine and dopamine agonists in the basal ganglia.