The treatment of hepatic encephalopathy is predominantly an empirical approach because the process pathogenesis of this syndrome is uncertain. Many treatment trials have been published, but few fullfil the usually accepted criteria for optimum study design. Lack of placebo controls, inadequate sample size and the difficulty in assessing the severity of encephalopathy are noteworthy examples. Treatments based on reducing ammoniagenic factors in the intestinal tract and promoting urinary waste nitrogen elimination are discussed. Other forms of therapy based on the false neurotransmitter and γ-aminobutyric hypotheses are reviewed before outlining a practical approach to treatment of hepatic encephalopathy in clinical practice.