Systemic hyperammonemia has been largely found in patients with cirrhosis and hepatic encephalopathy, and ammonia plays a major role in the pathogenesis of hepatic encephalopathy. However, controversial points remain: a) the correlation between plasma ammonia levels and neurophysiological impairment. The lack of correlation between ammonia levels and grade of hepatic encephalopathy in some cases has been considered a weakness of the ammonia hypothesis, but new methods for ammonia measurements and the implication of systemic inflammation in the modulation of ammonia neurotoxicity could explain this gap; b) the source of ammonia production. Hyperammonemia has been considered as derived from urea breakdown by intestinal bacteria and the majority of treatments were targeted against bacteria-derived ammonia from the colon. However, some data suggest an important role for small intestine ammonia production: 1) the hyperammonemia after porto-caval shunted rats has been found similar in germ-free than in non-germ-free animals. 2) In cirrhotic patients the greatest hyperammonemia was found in portal drained viscera and derived mainly from glutamine deamination. 3) The amount of time required to increase of ammonia (less than one hour) after oral glutamine challenge supports a small intestine origin of the hyperammonemia. As the main source of ammonia in cirrhotics derives from portal drained viscera owing to glutamine deamidation, increased glutaminase activity in the intestine seems to be responsible for systemic hyperammonemia. Lastly, some genetic alterations in the glutaminase gene such as the haplotype TACC could modulate intestinal ammonia production and the risk of overt hepatic encephalopathy in cirrhotics.