The purpose of this study was to determine the relationship between dopamine D₂ receptor occupancy and plasma haloperidol. Twelve patients treated with 1–5 mg/day of haloperidol had their D₂ occupancy measured using [¹¹C]-raclopride and positron emission tomography and haloperidol plasma levels measured using gas chromatograph mass spectrophotometer. The patients exhibited haloperidol plasma levels ranging from 0.5 to 5.8 ng/ml and D₂ occupancy from 53 to 88%. The D₂ occupancy was related to the plasma level as a saturating rectangular hyperbola relationship (r ² = 0.84) and it showed that, on average, 50% D₂ occupancy was achieved with 0.51 ng/ml and 80% D₂ occupancy with 2.0 ng/ml. Our findings demonstrate that 2–5 mg/day of haloperidol, which usually leads to plasma levels of 1–2 ng/ml, would be expected to induce 60–80% dopamine D₂ receptor occupancy. If, as has been claimed, 70% D₂ occupancy is adequate for typical neuroleptic response, then the conventional use of >10 mg/day may have been too high, since 70% occupancy can be achieved in most patients by 2–5 mg/day. On the other hand, if as others have suggested, 8–12 ng/ml of haloperidol is the correct therapeutic window for plasma levels, then the required therapeutic D₂ occupancy is closer to 90%, not 70%. The implications of the D₂ occupancy findings for the optimal dosing of neuroleptics are discussed.