A randomized double-blind trial of colchicine vs placebo was conducted in 67 patients with histologically proven alcoholic hepatitis, 33 of whom had cirrhosis. Patients with hepatic encephalopathy, ascites, protracted prothrombin time, severe thrombocytopenia, hepatocellular carcinoma, evident lack of discipline or refusal to participate in the trial were not included. Thirty-three patients received colchicine (1 mg/day) and 34 received placebo for 6 months. Blood parameters including N-terminal peptide of type III procollagen were assessed in the serum, and a percutaneous liver biopsy was performed at the start of the trial and after 3 and 6 months. Alcoholic hepatitis and fibrosis scores were established for each biopsy specimen. Twenty-eight percent of patients were lost to follow-up at 3 months, and fifty-two percent at 6 months. One patient died of liver failure. Fifty-eight percent of patients were abstaining from alcohol at 3 months and fifty percent at 6 months. No significant effect of treatment was noted. Nevertheless, improvement in alcoholic hepatitis core at 3 months was more important in the colchicine group than in the placebo group. No side-effects were noted except transient diarrhea. Our results suggest that colchicine has no important effect on the course of alcoholic hepatitis. A trial including of at least 260 patients might be necessary for the observed alcoholic hepatitis score difference at 3 months, favoring colchicine, to be statistically significant.