We conducted a placebo-controlled, double-blind study of valproate, a drug originally developed as an antiepileptic, in 36 patients with acute manic episodes who had previously failed to respond to or to tolerate lithium carbonate. Treatment duration was 7 to 21 days, with no other psychotropic medications permitted except lorazepam up to 4 mg/d during the first 10 days of treatment. Serum valproate concentrations were measured three times weekly; an unblinded investigator then adjusted dosage to produce serum concentrations between 50 and 100 mg/L. Valproate proved superior to placebo in alleviating manic symptoms. The 17 patients randomized to active drug demonstrated a median 54% decrease in scores on the Young Mania Rating Scale as compared with a median 5.0% decrease among the 19 patients receiving placebo. On the 100-point Global Assessment Scale of overall psychiatric functioning, patients receiving valproate improved by a median of 20 points as compared with a zero-point change with placebo. Significant differences also emerged on the Brief Psychiatric Rating Scale and in the number of supplemental doses of lorazepam required by the patients in each group. Substantial antimanic effects appeared within 1 to 4 days of achieving therapeutic serum valproate concentrations. Adverse effects were infrequent, with no adverse effect appearing significantly more frequently with valproate than with placebo. We conclude that valproate represents a useful new drug for the treatment of manic patients.