The specific D2 receptor agonist, LY 141865, but not the specific D1-receptor agonist, SK&F 38393, potently inhibited electrically evoked [3H]dopamine release from slices of the cat caudate. Similarly, LY 141865, but not SK&F 38393, inhibited [3H]spiperone binding to membranes of the cat caudate. The inhibition by dopamine receptor agonists of electrically evoked [3H]dopamine release was antagonized by the specific D2-receptor antagonist S-sulpiride. The inhibition of the electrically evoked release of [3H]dopamine by apomorphine was not, however, antagonized by the specific D1-receptor antagonist, bulbocapnine. Similarly, S-sulpiride but not bulbocapnine potently inhibited [3H]spiperone binding to membranes of the cat caudate. These results suggest that the dopamine autoreceptor modulating the depolarization-evoked release of [3H]dopamine, and the binding site of [3H]spiperone, are valid in vitro models for D2-dopamine receptors. Contrary to some previous reports, DPI was inactive in both in vitro dopamine receptor models. The IC50 values of a series of dopamine receptor agonists correlated very well in the two in vitro dopamine receptor models. One exception to this correlation was bromocriptine, which was more potent at [3H]spiperone binding sites than at the dopamine autoreceptor. With the exception of bromocriptine, all dopamine receptor agonists had one-hundred fold higher potency at the dopamine autoreceptor than at [3H]spiperone binding sites. [3H]Spiperone binding sites are localized primarily postsynaptic to dopamine terminals. Possible differences between the pharmacological properties of pre- and postsynaptic dopamine receptors should become apparent in the comparison of the two in vitro dopamine receptor models. However, the order of potency of dopamine receptor agonists with both in vitro models, dopamine autoreceptor and [3H]spiperone binding, was the same: N-n-propylnorapomorphine greater than TL-99 = 7-HAT greater than M-7 greater than Apomorphine greater than LY 141865.