Although the technique of endoscopic reflectance spectrophotometry has been applied in clinical studies, factors that modify the reproducibility of measurements have not been assessed systematically. To determine the limitations of the technique, measurements were made while endoscopic light intensity, systemic oxygen saturation, and orientation of the measuring probe were varied. The effects of hemorrhagic hypotension and exposure of the mucosa to 10% dextrose were also studied. When a large number (n = 480) of measurements in the human colon were considered, endoscopic light significantly decreased the index of oxygen saturation (ISO2) and increased the index of hemoglobin concentration (IHB). The decrease in ISO2, however, was small and unlikely to be of clinical importance despite being statistically significant. In one subject with chronic lung disease and baseline hypoxemia, administration of supplemental oxygen significantly increased oxygen saturation at the finger tip as measured by an oximeter and ISO2 of the buccal mucosa as measured by reflectance spectrophotometry. Varying the angle between the measuring probe and the gastric mucosa in rats from 90 degrees to 60 degrees did not affect ISO2 or IHB measurements. At 45 degrees, however, IHB but not ISO2 was significantly increased. Ischemia subsequent to induction of hemorrhagic hypotension and hyperemia induced by administration of 10% dextrose could be demonstrated reproducibly. We conclude that by lowering the intensity of endoscopic light and providing supplemental oxygen, errors in the measurement of IHB and ISO2, respectively, can be minimized. Minor deviations from the perpendicular orientation do not significantly affect ISO2 and IHB measurements. Attention to these details enhances the accuracy of endoscopic reflectance spectrophotometric recordings of ISO2 and IHB in clinical studies.