It has been argued that agonist ligands for dopamine D(2/3) receptors recognize a privileged subset of the receptors in living striatum, those which are functionally coupled to intracellular G-proteins. In support of this claim, the D(2/3) agonist [(3)H]-N-propylnorapomorphine ([(3)H]NPA) proved to be more vulnerable to competition from endogenous dopamine than was the antagonist ligand [(11)C]raclopride, measured ex vivo in mouse striatum, and subsequently in multi-tracer PET studies of analogous design. Based on these results, we predicted that prolonged dopamine depletion would result in a preferential increase in agonist binding, and a lesser competition from residual dopamine to the agonist binding. To test this hypothesis we used autoradiography to measure [(3)H]NPA and [(3)H]raclopride binding sites in hemi-parkinsonian rats with unilateral 6-OHDA lesions, with and without amphetamine challenge. Unilateral lesions were associated with a more distinct increase in [(3)H]NPA binding ex vivo than was seen for [(3)H]raclopride binding in vitro. Furthermore, this preferential asymmetry in [(3)H]NPA binding was more pronounced in amphetamine treated rats. We consequently predict that agonist ligands should likewise be fitter than antagonists for detecting responses to denervation in positron emission tomography studies of idiopathic Parkinson's disease. Agonist binding increases in vivo are likely to reflect the composite of a sensitization-like phenomenon, and relatively less competition from endogenous dopamine, as seen in the lesioned side of 6-OHDA induced hemi-parkinsonism.