In patients with hepatic encephalopathy (HE) the blood concentration of ammonia is usually highly elevated. Ammonia readily enters brain cells from the blood, and toxic effects of ammonia on brain metabolism and neurotransmission are believed to play a key role in the pathogenesis of HE. It has, however, been a matter of great controversy whether backflux of unmetabolized ammonia (NH(3) + NH(4) (+)) from brain cells to the blood occurs in man. In the present analysis of data from a dynamic PET study of brain (13)N-ammonia metabolism in healthy subjects and cirrhotic patients with and without HE, we provide the first unambiguous evidence for backflux of ammonia from brain cells to the blood in man. The high temporal and spatial resolution of modern PET technology was employed to distinguish between unidirectional blood-brain transport of ammonia and subsequent metabolism of the ammonia in the brain. In all 16 subjects, clearance of the unidirectional transport of (13)N-ammonia from the blood to brain cells (K(1)) was higher than the metabolic clearance of (13)N-ammonia from the blood (K(met)=K(1) k(3)/(k(2)+k(3)). This can only be explained by backflux (k(2)) of ammonia from brain cells to the blood. In conclusion, backflux of ammonia from the brain to the blood does indeed occur in both healthy subjects and cirrhotic patients with and without hepatic encephalopathy.