The purpose of this study was to determine the intracellular pH of the whole head and in voxels localized to the basal ganglia in patients with chronic liver disease using phosphorus-31 magnetic resonance spectroscopy (31P MRS). The study group compromised 82 patients with biopsy-proven cirrhosis (43 Child's grade A, 25 Child's grade B and 14 Child's grade C). Eleven subjects showed no evidence of neuropsychiatric impairment on clinical, psychometric and electrophysiological testing, 37 showed evidence of minimal hepatic encephalopathy and 34 had overt hepatic encephalopathy. Unlocalized 31P MRS of the whole head was performed in 48 patients and 10 healthy volunteers. Localized 31P MRS of the basal ganglia was performed in the 34 patients and in 20 healthy volunteers. The intracellular pH values were calculated from the chemical shift difference between the inorganic phosphate (P) and phosphocreatine (PCr) resonances. The percentage inorganic phosphate (%Pi), phosphocreatine (%PCr) and betaNTP signals, relative to the total 31P signal, and peak area ratios of inorganic phosphate and phosphocreatine, relative to betaNTP were also measured. There were no differences between patients and volunteers in intracellular pH in 31P MR spectra measured from the whole head or the basal ganglia. There was no correlation between the severity of encephalopathy (West Haven criteria) or liver dysfunction (Child score) and intracellular pH values. There was also no significant change in the inorganic phosphate, phosphocreatine or betaNTP resonances in spectra acquired from the whole head. However, in spectra localized to the basal ganglia, there was a significant increase in mean P/NTP (p=0.02) and PCr/NTP (p=0.009). The mean %Pi and mean %PCr were also increased (p=0.06; p=0.05, respectively), but there was no significant change in mean %betaNTP. When the patient population was classified according to the severity of encephalopathy, those with overt disease had a higher mean P/NTP and %Pi (p=0.03; p=0.01), compared to the reference population. Our results suggest that there are detectable bioenergetic abnormalities in patients with minimal hepatic encephalopathy or stable, overt chronic hepatic encephalopathy, but any associated intracellular pH change is probably a secondary, rather than a primary phenomenon.