Twenty-nine inpatients with major psychotic disorders were treated for 14 days with a clinician-determined dose of haloperidol and with either benztropine or placebo given by double-blind random assignment on days 1 through 7. No differences were noted in haloperidol mean dose, haloperidol blood levels, or BPRS scores during the first seven days between benztropine (N = 14) and placebo (N = 15) groups. Benztropine-treated patients demonstrated increased dry mouth and diminished sweat and a non-significantly lower rate of dystonia compared to placebo (14% vs. 33%). Dystonic patients were significantly younger than nondystonic patients, but did not differ in haloperidol mean dose or plasma concentration. The effect of benztropine on the incidence of dystonia was consistent with other studies, which, when analyzed together, demonstrate the efficacy of anticholinergic prophylaxis. The relatively low incidence of anticholinergic side effects, coupled with the lack of effect on haloperidol blood levels or antipsychotic efficacy, suggest that moderate doses of benztropine in conjunction with haloperidol are a rational approach for the treatment of acute psychosis in young patients.