The influence of psychologic factors on the healing and relapse of duodenal ulcers under treatment with ranitidine was studied in a prospective, multicenter trial in 2109 patients with an endoscopically proven duodenal ulcer (DU) and a history of recurrent duodenal ulceration. All patient received ranitidine (300 mg daily), and, after healing, 1899 patients continued maintenance treatment (ranitidine, 150 mg daily) for 2 years. A physician's assessment of stress (stress or no stress) was made at every consultation. In the healing phase an overall classification of stress as absent, intermittent, or continuous was made, and in the maintenance phase patients were classified dichotomously as having stress (stress on at least half of the follow-up consultations) or no stress. In addition, at the start of the healing phase stress was measured by means of a standardized questionnaire. Continuous stress, as assessed by the physicians, was associated with a lower 14-day healing rate (35.7%) than intermittent or absent stress (42.4%; relative risk (RR) for delayed healing in patients with continuous stress, 1.19; 95% confidence interval (CI), 1.06-1.33; P < 0.02). Differences in the 14-day healing rate for patients with low and moderate stress scores (43.1%) compared with those with high and very high stress scores (37.9%) just failed to reach statistical significance (RR for patients with stress, 1.14; 95% CI, 0.998-1.29; P = 0.051). During the 1st year of maintenance treatment 18.3% of patients with stress, but 10.9% of patients without stress, had a DU relapse (RR of stress for DU relapse during the first year, 1.73; 95 CI, 1.44-2.09; P < 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)