Substances that play an important role in the pathogenesis of hepatic encephalopathy (HE) are believed to originate in the gastrointestinal tract, particularly the colon, and to accumulate systemically in liver failure as a consequence of their impaired hepatic removal. Although the ideal goal of treatment for HE is to improve hepatocellular function, traditional therapies, such as lactulose and neomycin, are directed at reducing the systemic accumulation of comagenic substances by decreasing their intestinal absorption or synthesis. Thus these therapies act on the organ that is considered to be the origin of the substances causing the problem (1). An alternative approach is to give a treatment that acts on the target organ, the brain, by reversing neuropathophysiological events that directly contribute to the encephalopathy. That a benzodiazepine receptor (BZR) ligand might be a therapy of this type has been suggested by uncontrolled observations of ameliorations of HE in humans associated with the administration of the BZR antagonist, flumazenil (2-4).In this issue O f HEPATOLOGY, Steindl and her colleagues (5) explore the potential of three different BZR ligands in reversing manifestations of HE in an extensively characterized rat model of fulminant hepatic failure (FHF) (6,7). Is there a rationale for administering a BZR ligand in HE? Specifically, does a BZR-mediated abnormality of neuronal function contribute to HE, and could such an abnormality be amenable to reversal by a BZR ligand? To answer these questions it is first necessary to consider how the central BZR modulates neuronal function. This receptor is an integral component of the GABAA/BZR complex in synaptic neural membranes. The other components of the complex are a receptor for GABA (the GABAA receptor) and a chloride channel. After its release from presynaptic neurons, GABA binds to GABAA receptors on effector postsynaptic neurons. This binding triggers the opening of the chloride channel, which allows passage of chloride ions into the neuron and results in hyperpolarization of the surface membrane. These events are the basis of GABAmediated inhibitory neurotransmission. Gating of the chloride channel by the GABAA receptor is allosterically modulated by the BZR. The BZR may increase or decrease the efficiency of GABA-gated chloride conductance depending on the nature of ligands occupying the Address reprint requests to: E.