Aspirin and nonsteroidal antiinflammatory drugs have been implicated in the pathogenesis of gastrointestinal hemorrhage. To evaluate their impact on inpatients, charts from Temple University Hospital with a discharge ICD-9 code which included upper gastrointestinal hemorrhage during a one-year period were reviewed. Aspirin and/or nonaspirin nonsteroid antiinflammatory drug (NSAID) use was identified in 34 patients (19 daily users and 15 intermittent users). Sixty-seven patients who bled, but did not use these agents, served as controls. Daily NSAID users were older than intermittent users and controls (P < 0.05). A higher frequency of bleeding ulcers was associated with NSAID use. Patients using NSAIDs spent more time in intensive care than controls (median 1 day vs 0 days). Daily users had a higher transfusion requirement (4 units) than non-users (0 units; P < 0.05). This study suggests that NSAID use has a substantial impact on health care resource utilization in patients admitted to an urban hospital for upper gastrointestinal hemorrhage.