Twelve acutely manic patients, newly admitted from the emergency room, were treated wih clonazepam or lithium carbonate in a double-blind cross-over design: half the patients chosen randomly received 10 days of treatment with clonazepam followed immediately by 10 days of treatment with lithium, while the others received the same treatments in reverse order. Clonazepam was found to be significantly more efficacious than lithium in reducing manic symptoms despite the fact that during clonazepam treatment less patients required PRNs of haloperidol. Furthermore, both the total dose of PRN haloperidol and the number of days it was needed were significantly lower during clonazepam treatment than during lithium treatment. Clonazepam had a rapid onset of action, was highly sedative, and was well tolerated at high doses. By reducing the need for neuroleptics in the treatment of acute mania, clonazepam may decrease the risk of tardive dyskinesia in manic patients.