A controlled trial on the use of Silymarin in patients with alcoholic liver disease was performed. Seventy two patients were admitted to the trial and randomly assigned to an experimental or controls group. Experimentals received 280 mg/day of Silymarin and controls an equal number of placebo tablets. Three patients on placebo and nine on Silymarin were lost from control (p = 0.035), remaining in control 34 patients receiving placebo and 25 patients receiving the drug. Both groups did not differ in their initial laboratory assessment and were followed up for an average of 15 months. Ten patients died during the follow up (5 in placebo and 5 in Silymarin); life table analysis did not show significant differences in mortality. Patients who died had lower serum albumin and prothrombin time and higher total bilirubin, alkaline phosphatases and CCLI on the initial clinical and laboratory assessment. Final laboratory values and their changes in those who survived did not differ between Silymarin and placebo. Twenty two patients on placebo (65%) and 14 on Silymarin (58%) recognized alcohol ingestion or had a positive urine sample analysis for alcohol during follow up. Those who abstained from alcohol had a significant fall in gamma glutamyl transferase during follow up. No other significant differences were observed between these two groups. It is concluded that in this trial, Silymarin did not change the evolution or mortality of alcoholic liver disease.