Background and aim: This prospective randomized study was designed to assess the efficacy of 10-day and 14-day rifabutinbased triple therapy as a third-or fourth-line rescue therapy. Methods: Patients who failed first-and second-line eradication therapy were enrolled. H. pylori was isolated from gastric biopsy specimens and the rpoB mutation status, a factor of resistance to rifamycins, and minimum inhibitory concentrations (MICs) of rifabutin and amoxicillin were determined. Enrolled patients were randomly assigned to receive 10-day or 14-day eradication therapy with esomeprazole (20 mg, 4 times a day (q.i.d.)), amoxicillin (500 mg, q.i.d.), and rifabutin (300 mg, once a day (q.d.s.)). Poor compliance was defined as intake of <80% of study drugs. Successful H. pylori eradication was confirmed using a [13C] urea breath test or a stool antigen test, 12 weeks after the end of therapy. Results: Twelve patients were assigned to the 10-day group, and 17, to the 14-day group. Intention-to-treat and per-protocol analyses of eradication rates were 83.3% and 81.8% for the 10-day group and 94.1% and 91.7% for the 14-day group, respectively. All patients with rpoB mutation-positive strains (n ¼ 3) showed successful eradication, irrespective of the regimen received. Therapy was stopped due to adverse events in 8.3% and 29.3% of patients in the 10-day and 14-day groups, respectively. Conclusion: Both the 10-day and 14-day therapies were effective as rescue regimens. In particular, the 14-day therapy resulted in successful eradication in over 90% of patients, but the 10-day treatment may be enough to obtain a successful eradication rate, considering the tolerability of therapy.