Drug selection for the treatment of non-steroidal anti-inflammatory drug (NSAID)-induced gastric ulcer was analyzed pharmacoeconomically. Two patterns consisting of continuation of an NSAID plus administration of the prostaglandin (PG) preparation misoprostol (PG model) for 8 weeks and continuation of an NSAID plus administration of the proton-pump inhibitors omeprazole and lansoprazole (PPI model) for 8 weeks were examined. Decision analysis models were created on the basis of reports of clinical studies and epidemiologic studies relating to the drugs and gastric ulcer, and cost-comparative analyses were conducted based on the number of persons who had ulcer healing as health outcomes. Costs were estimated with respect to health expenditures from the third-party payer (public) perspective. In the case of continuation of an NSAID plus administration of the proton-pump inhibitor omeprazole for 8 weeks, the health outcomes improved and costs were reduced in comparison with continuation of an NSAID plus administration of misoprostol, thus making the administration of omeprazole the dominant choice. With continuation of an NSAID plus administration of lansoprazole for 8 weeks, the cost-savings of lansoprazole were inferior to those of misoprostol. The generic omeprazole product was the most cost-saving among the four drugs (misoprostol, original omeprazole product, generic omeprazole product, and lansoprazole) examined.