Although nonsteroidal antiinflammatory drug (NSAID) use is strongly associated with both upper gastrointestinal bleeding (UGIB) and lower gastrointestinal bleeding (LGIB), few data exist regarding the outcome of the bleeding episode for those consuming these drugs. Consecutive patients with UGIB or LGIB evaluated during the period August 1, 1990 through September 30, 1994 at a large inner city hospital were prospectively identified. Both prescription and over-the-counter NSAID use was specifically evaluated. Endoscopy was performed in most patients for diagnosis. Outcome measures included transfusion requirement, hospital stay, need for endoscopic therapy or surgery, and death. Over the 50-month study period, 785 patients admitted with UGIB and 161 with LGIB were studied. NSAID use was documented in 59% of patients with UGIB and 51% with LGIB. In UGIB, NSAID users were more likely to be female and older. NSAID users had a significantly shorter median hospital stay (4 vs 5 days), less rebleeding (11% vs 18%; P = 0.004) and in- hospital mortality (5% vs 13%: P = 0.001) as compared to nonusers. These differences remained significant when controlling for age, race, and gender. Similar trends in outcome were seen when evaluating ulcer- and non-ulcer- related bleeding. NSAID users with LGIB were more likely to be female, although rebleeding (19% vs 21%), hospital stay, and in-hospital mortality (5% vs 2%) were not significantly different between users and nonusers. UGIB in NSAID users appears to have a better prognosis as compared to nonusers. In contrast, NSAID use does not appear to be associated with outcome in patients with LGIB.