BACKGROUND AND AIMS: Genetic polymorphism of cytochrome P450 (CYP) 2C19 influences the efficacy of Helicobacter pylori eradication therapy with a proton pump inhibitor (PPI) and amoxicillin. However, in triple therapy (PPI plus amoxicillin and clarithromycin), little is known about the impact of CYP2C19 polymorphism, or the use of rabeprazole, which is not well metabolized by CYP2C19. The efficacy of three PPI (omeprazole, lansoprazole, and rabeprazole) in a 1-week triple regimen were compared in relation to CYP2C19 polymorphism. METHOD: One hundred and eighty-three patients were randomized to receive one of the following regimens: amoxicillin 500 mg t.i.d., clarithromycin 200 mg t.i.d., and PPI (omeprazole 20 mg, lansoprazole 30 mg, or rabeprazole 10 mg) b.i.d. CYP2C19 polymorphism was analyzed by PCR restriction fragment length polymorphism. RESULTS: Intention-to-treat-based overall cure rates for omeprazole, lansoprazole or rabeprazole regimens were 83.1% (95% confidence interval (CI): 69-89%), 86.7% (CI: 75-93%), and 76.6% (CI: 64-85%), respectively, without significant difference. The cure rate of the rabeprazole regimen (but not the lansoprazole or omeprazole regimens) tended to be correlated with CYP2C19 genotypes (P = 0.076). In patients with a homozygous extensive metabolizer genotype, the per protocol-based cure rate with rabeprazole (62.5%) was significantly lower than that with lansoprazole (90.0%; P = 0.038). CONCLUSION: The overall cure rate of 1-week triple therapy for H. pylori eradication was not significantly different between regimens with omeprazole, lansoprazole or rabeprazole, but the impact of CYP2C19 genetic polymorphism on the cure rate appeared to differ between these PPI.