Hospitalized patients with hepatic insufficiency often suffer from severe catabolic states and are in urgent need of nutritional support during their acute illness. Protein intolerence, however, remains a significant problem with respect to the provision of adequate nutrition, either enterally or parenterally. The following report is an anecdotal series of 63 consecutive patients in a large urban hospital treated prospectively with nutritional support using a prototype high branched-chain amino acid solution (FO80) given by technique of total parenteral nutrition by the subclavian or internal jugular route with hypertonic dextrose. Sixty-three patients, of which 42 had chronic liver disease (cirrhosis) with acute decompensation and 17 with acute hepatic injury as well as four with hepatorenal syndrome, are the subject of this report. All required intravenous nutritional support and were either intolerant to commercially available parenteral nutrition solutions or were in hepatic encephalopathy at the time they were initially seen. The cirrhotic patients had been hospitalized for a mean of 14.5 +/- 1.9 days before therapy, had a mean bilirubin of 13 mg/100 ml, and had been in coma for 4.8 +/- 0.7 days despite standard therapy. Patients with acute hepatitis had been in the hospital for 16.2 +/- 4.1 days before therapy, had a mean bilirubin of 25 mg/100 ml, and had been in coma 5.2 +/- 1.6 days before therapy. Routine tests of liver function, blood chemistries, amino acids, EEGs, and complex neurological testing including Reitan trailmaking tests were used in the evaluation of these patients. Up to 120 grams of synthetic amino acid solution with hypertonic dextrose was tolerated in these patients with improvement noted in encephalopathy of at least one grade in 87% of the patients with cirrhosis and 75% of the patients with hepatitis. Nitrogen balance was achieved when 75 to 80 grams of synthetic amino acids were administered. Survival was 45% in the cirrhotic group and 47% in the acute hepatitis group. Encephalopathy appeared to correlate with individual amino acids differentially in the various groups and with the ratio between the aromatic and the branched-chain amino acids. Ammonia did not correlate with either the degree of encephalopathy or improvement therefrom. In 24 Patients therapy for hepatic encephalopathy was limited to infusion of the branched-chain enriched amino acid solution only, with wake-up in 66% of this group. The results strongly suggest that in protein intolerant patients requiring nutritional support, infusion with branchedchain enriched amino acid solutions is well tolerated with either no worsening of or improvement in hepatic encephalopathy coincident with the achievement of nitrogen equilibrium and adequate nutritional support.