In this study, the authors examined the relationship between steady-state haloperidol blood levels and clinical response in patients with acute psychotic mania. Fifty-four inpatients with acute mania were randomly assigned to receive either haloperidol 25 mg/day or haloperidol 5 mg/day. Each subject also received a concomitant medication: lorazepam 4 mg/day, lithium, or placebo. The relationship between steady-state haloperidol blood levels and clinical improvement was studied using analysis of covariance. There was wide interindividual variation in the haloperidol blood level-dose ratio. Haloperidol blood levels (log-transformed) were found to significantly correlate with clinical response in acute mania. Low-dose haloperidol with concomitant lithium may produce an optimal response in acute mania. Haloperidol blood levels may be clinically useful in identifying patients who are nonresponsive because of low drug levels and, hence, in enhancing optimal haloperidol dosing for acute mania with psychosis.