In a controlled cross-over trial, we have compared a conventional 40-g protein diet (30 g animal and 10 g vegetable, diet A) with an 80-g vegetable-protein-supplemented diet (30 g animal and 50 g vegetable, diet B) in the treatment of six patients with chronic stable portal systemic encephalopathy, requiring dietary and lactulose therapy. Each diet was given, in random order, for 5 days in hospital. EEG, clinical indices of encephalopathy, and the plasma amino acid profile were assessed at the end of each treatment period. The increase in vegetable protein intake was associated with minor improvement in EEG and clinical performance in two patients, and no change in the others. Fasting plasma phenylalanine and tyrosine were higher on diet B [phenylalanine 108.6 +/- 9.3 (SEM) mumol/L versus 99.6 +/- 8.37, p less than 0.05 (paired t test); tyrosine 153 +/- 15.2 mumol/L versus 140 +/- 14, p less than 0.05). The plasma branched-chain amino acid levels did not change, and the branched chain/aromatic amino acid ratio (BCAA/AAA) was lower on diet B (p less than 0.02). Fecal weights were not significantly altered. These results indicate that patients with chronic portal systemic encephalopathy are tolerant of protein supplementation from vegetable sources. A minor improvement in parameters of encephalopathy was seen in some individuals, despite a lowering of BCAA/AAA which some investigators have thought important in the pathogenesis of encephalopathy.