We set up a new system for measuring the gastroduodenal HCO3- responses using pH deflection and the potential difference (PD) in the anesthetized rat. The stomach or the proximal duodenum was perfused at a flow rate of 0.7 ml/min with saline (pH 4.5), the pH of the perfusate and PD were continuously monitored, and HCO3- output was determined as acid-neutralizing capacity by back-titration of the perfusate to pH 4.5. In the case of the stomach, acid secretion was inhibited by omeprazole (60 mg/kg i.p.). Under these conditions, the pH, PD, and HCO3- output were 5.5 +/- 0.03, -3.8 +/- 0.4 mV, and 1.5 +/- 0.3 microEq/10 min in the duodenum and 5.4 +/- 0.04, -52.4 +/- 1.8 mV, and 1.2 +/- 0.3 microEq/10 min in the stomach, respectively. In addition, when various amounts of NaHCO3 were added into the system, a linear relationship was obtained between the area of pH deflection and the amount of added HCO3- (r = 0.999). Both pH and HCO3- output in these tissues were significantly increased by intravenous administration of prostaglandin E2, carbachol, and YM-14673 (a TRH analogue); the net HCO3- output in the duodenum was 8.7 +/- 1.3, 2.3 +/- 0.5, and 5.2 +/- 0.9 microEq, respectively. The values of net HCO3- output measured by back-titration coincided well with those obtained from the area of pH deflection caused by various agents. These results indicate that this system using pH deflection may be useful for quantitative determination of HCO3- response in the gastroduodenal mucosa.