Rheumatoid arthritis (RA) patients are at increased risk of peptic ulcers (PU) induced by nonsteroidal antiinflammatory drugs (NSAIDs). However, the impact of potential drug interactions on the development of PU has yet to be determined in a daily clinical setting. The aim was to estimate the clinical important interactions for PU presented by comedication in Japanese RA outpatients on long-term NSAID treatment. This retrospective cohort study enrolled 196 consecutive RA outpatients on NSAID medication for at least 3 months. Potential risk factors for endoscopic PU were analyzed in RA outpatients on longterm NSAID treatment. PU incidence was 31% with bisphosphonate co-therapy and 17% without the co-therapy. PU incidence was only 5% in subjects with proton pump inhibitors (PPI) or prostaglandin E1 analogues (PG) co-therapy, 14% with histamine-H(2) receptor antagonists(H2RA) co-therapy, and 27% without anti-ulcer agents. In multivariate logistic regression analysis, bisphosphonate co-therapy remained a significant risk factor for PU (OR, 2.29; 95% CI, 1.09-4.81). Other risk factors for ulcer development were advanced age (greater than 60 years) and smoking (OR, 2.58; 95% CI, 1.03-6.49 and OR, 2.71; 95% CI, 1.13-5.53, respectively.) Factors that significantly reduced the incidence of PU were H2RA or PPI/PG cotherapies (OR, 0.29; 95% CI, 0.12-0.68.). Bisphosphonate co-therapy as well as advanced age and smoking was found to be a significant risk factor in PU, while co-therapies of standard-dose H2RA or PPI/PG proved effective in preventing PU in Japanese RA patients on long-term NSAID treatment.