The aim of this study was to evaluate whether PCR-based restriction fragment length polymorphism (RFLP) analysis was effective in differentiating between reinfection and recrudescence of H. pylori strains. Following a 1-2 week regimen of omeprazole 20 mg, amoxicillin 1.0 g, and clarithromycin 500 mg twice daily, twenty patients with duodenal ulcer were enrolled in the study. Ten patients (group 1, control) were not successfully treated, and another 10 patients (group 2) exhibited recurrence of infection 6-24 months following the therapy. Follow-up diagnosis was performed by Giemsa stain and CLO test. RFLP profiles of antral and midbody biopsy specimens were compared before and after therapy. PCR products using the ureC gene were digested with restriction enzymes Hha I, Mbo I, and Hind III, and the fragments generated were analyzed by agarose gel electrophoresis. Hha I, Mbo I, and Hind III digestion produced 13, 7, and 2 distinguishable digestion patterns, respectively. There was no difference in RFLP profiles seen before and after the therapy in 17 duodenal ulcer patients, while different RFLP profiles were discovered in 3 patients. Following treatment, one (group 2) patient differed in Mbo I, and two (one each from both groups) patients differed in Hha I and Mbo I RFLP patterns. Eight of group 2 patients showed recrudescence of previous infection and two patients had reinfection by another strain. This study supports the hypothesis that PCR-based RFLP analysis can be effective for differentiating reinfection and recrudescence of H. pylori strains following triple therapy.