A randomised double-blind trial of thioctic acid (alpha-lipoic acid), 300 mg/day versus placebo was carried out in 40 patients with pre-cirrhotic alcohol-related liver disease over a six month period. Twenty patients received the active drug and 20 placebo. Twenty-two of the 40 patients (55%) abstained from alcohol and showed significant improvements (p less than 0.01) in mean values for serum aspartate transaminase, serum glutamyl transpeptidase, and mean corpuscular volume. Seventeen of the 22 (77%) showed overall histological improvement on liver biopsy. The remaining 18 patients (45%) continued to drink but significantly reduced their mean daily alcohol intake (p less than 0.001). No significant changes occurred in their laboratory indices, but five of the 18 (28%) showed overall histological improvement. Changes occurred irrespective of treatment with thioctic acid, which suggested that, over six months, this drug did not influence the course of alcohol-related liver disease.