Nonsteroidal antiinflammatory drug-induced gastroduodenal mucosal damage observed endoscopically is usually categorized as hemorrhages, erosions, or ulcerations. We undertook this study to determine whether the injury produced by a commonly prescribed NSAID, naproxen, could be reduced by cotherapy with sucralfate or cimetidine and to determine how dependent the differences in the degree of protection against mucosal injury measured were on the scoring system used. Four groups of 20 healthy volunteers with endoscopically normal gastric and duodenal mucosa received naproxen (500 mg twice a day) plus cimetidine (300 mg four times a day or 400 mg twice a day), sucralfate (1 g four times a day), or placebo for seven days. After seven days of therapy, a second endoscopy was performed. Separate scoring systems were used for the presence of hemorrhages, erosions, and a combination of both types of injury. There were significantly fewer mucosal hemorrhages present when naproxen and cimetidine were administered than when naproxen was administered with placebo or sucralfate (placebo vs 300 mg cimetidine, P = 0.04, and placebo vs 400 mg cimetidine, P = 0.006, placebo vs sucralfate, P = 0.26). Both cimetidine dosages resulted in significantly fewer hemorrhages than were present following cotherapy of naproxen and sucralfate (P less than 0.05). In contrast, there was no discernible difference in the mucosal injury between placebo and any drug or between any two active therapies when the injury was evaluated based on the presence of gastric erosions.(ABSTRACT TRUNCATED AT 250 WORDS)