In quantitative studies of gastric functions, calculations based on the recovery of an ideal non-absorbable reference substance from the stomach allow corrections for the loss of the gastric contents through the pylorus. When the gastric contents of histamine stimulated rats were aspirated 45 minutes after oral administration of test meals at pH 6.5 containing phenol red or 51Cr-EDTA, the mean recovery of phenol red (37.0 % f 1.4) was significantly lower than the corresponding value for 51Cr-EDTA (48.7 % f 2.2). Consequently, when the calculations were based on phenol red as a reference substance, significantly larger volumes of fluid and higher amounts of HCI were estimated to have passed through the pylorus. In a paired comparison with the two compounds given in the same test meal, similar results were obtained. Mucosal binding and absorption of phenol red are suggested as the most likely explanations for the discrepancy. In addition to high stability, inertness, and a more complete recovery, the water-soluble 51Cr-EDTA complex offers analytical advantages. In the present method 51Cr-EDTA thus seems to meet the requirements of a satisfactory reference substance better than phenol red.