This study assesses the effect of oral iron therapy on the upper gastrointestinal tract and fecal occult blood testing. Fourteen healthy volunteers completed a checklist of gastrointestinal symptoms, underwent endoscopy and biopsy of the stomach and duodenum, and supplied a fresh stool sample for Hemoccult and HemoQuant testing. They then took ferrous sulfate 325 mg per os tid for two weeks and had the same evaluation repeated. Gastrointestinal symptoms were rated by the patients on a scale of 0-3, endoscopic findings were numerically scored (0-4), and the biopsies were graded blindly. Thirteen other healthy volunteers took ferrous sulfate 325 mg per os tid for one week and had Hemoccult testing of stool at days 0 and 7. All subjects developed dark stools, and significant nausea and diarrhea were noted (0.1 +/- 0.1 to 0.9 +/- 0.3, P less than 0.05 for both symptoms). Only 1/27 had a questionably trace-positive Hemoccult test (two observers disagreed) and no significant difference was seen in HemoQuant testing (1.4 +/- 0.5 to 1.8 +/- 0.7 mg Hb/g). A significant increase was seen in endoscopic abnormalities in the stomach (0.1 +/- 0.1 to 1.5 +/- 0.3, P = 0.003), consisting of erythema, small areas of subepithelial hemorrhage, and, in two subjects, erosions. Biopsies showed no significant change after iron therapy. We conclude that (1) oral ferrous sulfate rarely causes Hemoccult-positive stools, and patients with positive Hemoccult tests on iron therapy require further evaluation; and (2) oral iron may cause mild endoscopic abnormalities in the stomach which are of uncertain clinical significance.