The effects of bromfenac sodium, aspirin, and placebo on gastrointestinal (GI) blood loss were compared. In a 22-day, randomized study, healthy men received treatment with either bromfenac sodium 300 mg/d, aspirin 3900 mg/d, or placebo for 10 days. On days 3 through 9 and days 20 through 22, all patients received placebo. Fecal blood was measured using the chromium 51-labeled red blood cell technique. Thirty-seven subjects entered the treatment period (13 in the aspirin group, 12 in the bromfenac sodium group, and 12 in the placebo group). The mean change in fecal blood loss during the treatment period compared with the baseline period was significantly greater in the aspirin group (8.00 +/- 4.17 mL/d) than in the bromfenac sodium group (1.63 +/- 1.01 mL/d). Blood loss in both the aspirin and bromfenac sodium groups was significantly greater than in the placebo group (-0.12 +/- 0.25 mL/d). It is concluded that bromfenac sodium 300 mg/d, a higher daily dose than the proposed daily dose, causes significantly less GI blood loss than aspirin 3900 mg/d.