Acute gastroduodenal injury is commonly associated with the use of nonsteroidal anti-inflammatory drugs. The mechanism of injury is not well understood. The objectives of this study were to evaluate the protective effect of two drugs that give different degrees of acid inhibition against naproxen-induced gastroduodenal injury. Fifteen volunteers aged 22-28 years underwent pre- and post-treatment gastroduodenoscopies during three treatment periods (that is, six examinations), and mucosal injury was graded on a Lanza scale ranging from 0 to 4. The subjects received placebo, 150 mg rantidine twice daily, or 40 mg omeprazole in a double-blind, random-order design for 7 days. Plain naproxen, 500 mg twice daily, was given on days 3-7. The mean injury score for the stomach during placebo treatment was 1.53, and ranitidine gave 44% and omeprazole 40% reduction compared with placebo, which did not reach statistical significance. About 70% of the stomach injury was located in the antrum. The mean injury score during placebo for the duodenum was 1.93, and ranitidine gave 80% and omeprazole 90% reduction (p = 0.004). In conclusion, a correlation between different degrees of acid suppression and a protective effect on the gastroduodenal mucosa could not be shown. The study suggests that acid plays a major role in acute naproxen-induced injury to the duodenal mucosa, and a moderate acid reduction is adequate for protection. In the stomach acid seems to play a minor role in the mucosal injury, but physiochemical contact with naproxen in the antrum and a cyclooxygenase inhibition are of greater importance.