Benzodiazepines are the drugs of choice in the treatment of alcohol withdrawal syndrome (AWS). Recent data have shown that baclofen may reduce AWS symptoms. At present, no comparative studies between baclofen and any benzodiazepine used in AWS treatment are available. Accordingly, the present study was designed to compare efficacy, tolerability and safety of baclofen versus diazepam in the treatment of AWS.Thirty-seven patients with AWS were enrolled in the study and randomly divided into 2 groups. Baclofen (30 mg/day for 10 consecutive days) was orally administered to 18 patients (15 males, 3 females; median age: 46.5 years). Diazepam (0.5-0.75 mg/kg/day for 6 consecutive days, tapering the dose by 25% daily from day 7 to day 10) was orally administered to 19 patients (17 men, 2 women; median age: 42.0 years). The Clinical Institute Withdrawal Assessment (CIWA-Ar) was used to evaluate physical symptoms of AWS.Both baclofen and diazepam significantly decreased CIWA-Ar score, without significant differences between the 2 treatments. When CIWA-Ar subscales for sweating, tremors, anxiety and agitation were evaluated singly, treatment with baclofen and diazepam resulted in a significant decrease in sweating, tremors and anxiety score, without significant differences between the 2 drug treatments. Both treatments decreased the agitation score, although diazepam was slightly more rapid than baclofen.The efficacy of baclofen in treatment of uncomplicated AWS is comparable to that of the "gold standard" diazepam. These results suggest that baclofen may be considered as a new drug for treatment of uncomplicated AWS.