Both hyperammonemia and blood-brain barrier (BBB) breakdown have been implicated in the evolution of hepatic encephalopathy. To define a possible relationship, Swiss Albino mice were subjected to sublethal encephalopathic doses of ammonium acetate; the integrity of the BBB was determined grossly with Evans blue and quantitatively with [ 14C]-c-aminoisobutyratc: (AIB). Some animals were injected with a dose of ammonium acetate sufficient to maintain coma for 1 hr (AC group). One group, termed stuporous (AS), received only enough ammonium acetate to interfere with grooming and exploratory activity; this dosage was insufficient to completely block the righting response, which was absent in the AC group. When compared to that of controls (CON) receiving normal saline instead of ammonium acetate, cerebral tisue from the AC group was stained blue and contained nearly double the amount of AIB; AS group brain tissue was unstained and the AIB content did not differ significantly from normal. Some of the AC group were pretreated with drugs known to retard BBB breakdown; one set received dexamethasone (AC-DXMN), another the ornithine decarboxylase inhibitor difluoromethyl ornithine (AC-DFMO), and a third L-ornithine (AC-ORN). Brain tissue from the AC-ORN group stained blue and AIB content did not differ significantly from that of the untreated AC group. Cerebral tissue of the AC-DXMN pretreatment group stained light blue; AIB content was significantly lower than in the AC group and greater than the CON group. 'The AC-DFMO brains were unstained and AIB content was significantly lower than in the AC group but did not differ significantly from CON. These results indicate that hyperammoneniia may induce BBB breakdown but that the disruption of barrier integrity is not antecedent to the development of coma, although it sccms to coincide with coma in time. DFMO is more effective than DXMN in preventing ammonium acetateinduced BBB breakdown, but in this study neither drug prevented the development of coma.