OBJECTIVE: Subjective effects of alcohol in alcoholics treated with naltrexone or placebo were compared. METHOD: In a previously reported double-blind clinical trial of 50 mg/day of naltrexone or placebo for treatment of alcoholism, 36 of 70 detoxified male veterans deviated from abstinence. Of these 36, 29 subsequently reported on the subjective effects of drinking during the trial. RESULTS: A larger proportion of naltrexone-treated subjects (seven of 12) than placebo- treated subjects (two of 17) reported that the "high" produced by alcohol during the study was significantly less than usual. The naltrexone-treated subjects also drank less alcohol than the placebo- treated subjects during the first drinking episode. There was no difference between groups in reported intoxication, craving, memory, or loss of temper. CONCLUSIONS: The lower alcohol consumption by the naltrexone-treated subjects may have resulted from naltrexone's blockage of the pleasure produced by alcohol.