Positron emission tomography (PET) antagonist ligands such as [(11)C]-raclopride are commonly used to study dopamine D2 receptor (D2) binding of antipsychotics. It has been suggested that agonist radioligands bind preferentially to the high-affinity state of D2 receptor and may provide a more relevant means of assessing D2 occupancy. The main objective of this study was to determine if D2 receptor occupancy (RO) could be differentiated with agonist and antagonist radioligands in vivo. Agonist radioligands [(3)H]-MNPA and [(3)H]-(+)-PHNO were synthesized and compared to antagonist [(3)H]-raclopride in the in vitro binding and in vivo occupancy studies. In vivo, unanesthetized rats were pretreated with quinpirole (full agonist), aripiprazole (partial agonist), or haloperidol (antagonist) prior to administration of the agonist or antagonist radioligand. All three pretreatment compounds showed equivalent dose-dependent D2 receptor occupancy in the rat striatum with each radioligand. The in vivo receptor occupancy results suggested that the binding of quinpirole, aripiprazole, and haloperidol to the high or low affinity state of the D2 receptor could not be differentiated using radiolabeled agonists or antagonists, presumably due to a predominance of high affinity states of the D2 receptor in vivo. This hypothesis was supported in part by the in vitro binding results. Our in vitro results show that [(3)H]-MNPA binds to D2S transfected CHO cell membranes at a single high affinity site. Displacement of [(3)H]-(+)-PHNO binding by quinpirole and elimination of most [(3)H]-(+)-PHNO binding by the guanine nucleotide GppNHp in striatal membranes suggest that the majority of D2 in striatal tissue is G-protein coupled. Together, these findings suggest that D2 agonist radioligands produce in vivo receptor occupancy comparable to [(3)H]-raclopride.