Omeprazole is mainly metabolized in the liver by CYP2C19, a genetically determined enzyme, while rabeprazole is mainly nonenzymatically degraded with a minor involvement by CYP2C19. We investigated the gastric ulcer healing effect of omeprazole versus rabeprazole evaluated endoscopically with reference to the different CYP2C19 genotypes. Eighty patients with active gastric ulcer were treated with a daily dose of 20 mg of omeprazole or 10 mg of rabeprazole. The endoscopic evaluation was performed at the baseline and 2- and 8-week posttreatment periods. The endoscopic improvement of gastric ulcer size and ulcer healing rates using a thin rubber disc with a diameter of 6 mm, were evaluated in relation to the CYP2C19 genotypic status. The mean 2-week posttreatment ulcer size value by rabeprazole did not significantly differ among the different CYP2C19 genotypes, whereas the mean value in the homozygous extensive metabolizer patients treated with omeprazole was significantly (P = 0.0057) greater than in those with rabeprazole. However, after the 8-week treatment, omeprazole and rabeprazole showed the similarly high healing rates of 87.8% (31/37) and 88.9% (32/36), respectively. Although both omeprazole and rabeprazole showed a high healing rate of gastric ulcer after the 8-week treatment period, the healing effect of rabeprazole appears to be relatively independent of the CYP2C19 status, resulting in an earlier repair of gastric mucosal damage evaluated endoscopically compared to that of omeprazole.