In a prospective study from August to the end of December 1992, all patients with endoscopically confirmed gastric or duodenal ulcers were examined for the presence of H. pylori infection by means of an antibody blood test and gastric biopsy. After a detailed history had been taken the patients were divided into two groups. Group I (n = 40) was made up of patients taking NSAIDs. Group II (n = 42) consisted of non-NSAID patients. A control group was formed of NSAID patients with no marked intestinal complaints and the absence of ulcers confirmed on endoscopy (group III, n = 38). H. pylori infection was identified in 33 out of 40 patients in group I (82.5%), 37 out of 42 patients in group II (88.0%), and only 17 out of 38 cases in group III (44.7%). Bleeding complications were evidently associated with the use of NSAIDs: 12 patients in group I (30%), as opposed to only 2 patients in group II (5%) showed bleeding from an ulcer on endoscopy. It would, thus, appear that ulcers frequently arise as a result of H. pylori infection, whereas bleeding complications of ulcers are associated mainly with NSAID intake. The QuickVue antibody test and biopsy results concurred in 81.8% of cases. Because of its high sensitivity the QuickVue test could become a very important risk assessment tool when planning long-term NSAID therapy. H. pylori eradication ought perhaps be considered before the start of NSAID treatment.