The purpose of this study was to examine whether blockade of either dopamine D1-like or D2-like receptors by selective antagonist administration into the nucleus accumbens (NAc) is sufficient to reverse cocaine-induced locomotion, and to develop a new technique that enables the population of receptors occupied by the antagonists to be quantified. Locomotor activity was assessed in rats that had received bilateral intra-accumbens injections of the D1-selective antagonist SCH-23390 (0–3.0 μg/0.5 μl/side) or the D2/D3-selective antagonist sulpiride (0–1.0 μg/0.5 μl/side), followed 15 min later by injections of saline or cocaine (15 mg/kg, i.p.). To assess receptor occupancy by the antagonists, 105 min prior to sacrifice the rats received intra-accumbens injections of the antagonist, followed 15 min later by an injection of the non-selective irreversible antagonist, N-ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline (EEDQ; 10 mg/kg, i.p.). Receptors were labeled with [ 3 H]SCH-23390 or [ 3 H]sulpiride in sections containing the NAc, and the autoradiograms allowed quantitation of receptors occupied (i.e. protected from EEDQ) by the antagonist given in vivo. Only a dose of 3 μg/side SCH-23390 reversed cocaine-induced locomotion, whereas a dose of 0.5 μg/side did not alter cocaine-induced locomotion despite occupying the same amount of [ 3 H]SCH-23390 binding sites in the NAc. Intermediate doses of 0.1 and 0.3 μg/side sulpiride reversed cocaine-induced locomotion, and also occupied the greatest number of [ 3 H]sulpiride binding sites in the NAc. The results suggest that blockade of D2-like, but not D1-like, receptors in the NAc is sufficient to reverse cocaine-induced locomotion, and also demonstrate the importance of quantifying receptors occupied by drugs administered intracranially.