A total of 478 patients with endoscopically confirmed duodenal ulcer entered this randomized, parallel, double-blind trial. Patients were randomly assigned to receive ebrotidine (N-[(E)-[[2-[[[2-[(diaminomethylene)amino]- 4-thiazolyl]methyl]thio]ethyl]amino]methylene]-4-bromo-benzenesulfona mid e, CAS 100981-43-9, FI-3542) 400 mg or ranitidine 300 mg tablets (4:1) respectively, administered in single evening doses. Endoscopy, clinical examination and symptom assessment were performed at baseline and at weeks 4 and 8. Safety evaluations including laboratory tests, treatment compliance and antacid consumption checks were conducted at the beginning and/or at the 4 and 8 week visits. Patients whose ulcer showed endoscopic healing at the 4-week control left the study. Both groups were matched in all parameters studied. The healing rates at 4 weeks were 76.4% and 75.3% for ebrotidine and ranitidine respectively, while at 8 weeks the final rates were 95% and 91.8% respectively. Accompanying symptoms disappeared rapidly and the patients returned to normal. Smoking proved to be a highly significant negative risk factor, since healing rates were 83.4% and 71.2% at 4 weeks and 97.4% and 92.3% at 8 weeks in non-smokers and smokers respectively (p = 0.0046). Smokers treated with ranitidine showed significantly lower final healing rates than non-smokers (86% vs 100%; p = 0.0358), while the healing rates among patients treated with ebrotidine were similar regardless of whether they were smokers or not (93.9% and 96.7% N.S.). Ebrotidine (94%) proved to be more effective than ranitidine (86%) in smokers with higher healing rates (p < 0.05). Alcohol intake showed no significant relationship with the healing rates. Both drugs demonstrated an excellent safety. There were no changes in blood parameters, and no significant adverse events were reported.