The Acute Hepatic Failure Study Group (AHFSG) has conducted a double-blinded, randomized evaluation of hydrocortisone in patients with acute hepatic failure. From July 1975 through August 1978, a 38-month period, 18 medical centers in the United States and one in Canada participated in this trial. A total of 64 patients were accessed and found eligible to participate in the study; two of them were subsequently eliminated from our analysis. Eighteen patients received placebo; 23 received 400 mg hydrocortisone per day, and 21 patients were administered 800 mg hydrocortisone per day. We did not observe any therapeutic effect of hydrocortisone, and the survival rates for placebo versus 400 mg and versus 800 mg hydrocortisone per day were 22%, 9%, and 24%, respectively. Fulminant hepatitis associated with drug hepatotoxicity or non-A, non-B hepatitis seemed to have a worse prognosis than fulminant B, although these differences were not significant. Serum alpha-fetoprotein had a modest prognostic value of survival and seemed to be limited to fulminant B. The AHFSG recommends, therefore, that corticosteroid use in acute hepatic failure with hepatic encephalopathy be discontinued.