Hepatic encephalopathy due to cirrhosis is frequently precipitated by exogenous factors, and the effectiveness of a specific treatment with neomycin sulfate has so far not been submitted to clinical trials. Over a period of five years, 102 cirrhotic patients developed hepatic encephalopathy at admission or during hospitalization, and 39 were randomized for treatment with either neomycin sulfate or placebo. Exclusion criteria were: 1. current usage of specific treatment for hepatic encephalopathy, 2. chronic hepatic encephalopathy and 3. multiple organ failure syndrome associated with hepatic encephalopathy. The group of excluded patients (n = 63) was compared with the randomized group (n = 39), and no statistical differences were found regarding sex and age distributions, Child-Pugh classification, etiology of cirrhosis, percipitating factors and grade of hepatic encephalopathy. These same parameters were also comparable among the 20 patients who received active neomycin and the 19 who were treated with placebo. The therapy for hepatic encephalopathy consisted in the control of precipitating factors associated with 6 g of neomycin sulfate "per os" or placebo. Therapeutic failure and death by the fifth day of treatment, occurred in four patients (10.2%), two in each of the randomized groups. The time elapsed between the initiation of the therapeutic procedure and regression to grade zero of hepatic encephalopathy was 39.11 +/- 23.04 hours for the group of active neomycin, and 49.47 +/- 21.92 hours for the placebo group, but this difference did not achieve statistical significance.