Factors that affect Pco(2) measurement in balloon saline during gastrointestinal tonometry are unclear. They include carbon dioxide diffusion rate, correction factors for calculation of equilibrium Pco(2) from measurements at saline dwell times that are shorter than needed for full equilibration, role of blood-gas analyser bias during ex vivo Pco(2) measurements in saline, and normal values for intragastric Pco(2) (Pi(co2)) and intramucosal pH (pHi) at equilibrium, and their differences from blood values. In a laboratory study, normal Pco(2) changes in a saline-filled tonometer balloon placed in a saline bath at constant Pco(2) were described by a non-linear model, with a half-time of mean 4.4 min and 95% equilibration at mean 83 min. In a study in 20 healthy volunteers, A,? build up in a saline-filled tonometer balloon placed in the stomach, measured at dwell times of 10, 20, 30 and 60 min, was slightly (P < 0.05) slower than in vitro, with a half-time of mean 5.8 min and 95% equilibration at mean 110 min. Correction factors to derive equilibrium Pico, at short dwell times and independently from blood-gas analyser bias were calculated. The factors differed (P < 0.05) from those currently provided by the manufacturer. Normal threshold values (mean) were: equilibrium Pi(co2) less than or equal to 6.6 kPa, pHi greater than or equal to 7.33, Pi(co2) blood Pco(2) difference less than or equal to 1.1 kPa and pH difference greater than or equal to-0.06. Pi(co2) did not differ from, and was directly related to, blood P-co2. These values provide a reference base for other studies and show that gastric mucosal Pco(2). depends on alveolar ventilation if blood flow is adequate.