OBJECTIVES: The prevalence of bleeding from reflux esophagitis has not been studied. The aim of the study was to evaluate the 1-yr prevalence of bleeding from reflux esophagitis, as well as the independent factors associated with bleeding. METHODS: All patients with reflux esophagitis diagnosed with upper digestive tract endoscopy in Reims Hospital in 1996 were included. Studied parameters were prospectively recorded and compared between patients with bleeding and nonbleeding reflux esophagitis. RESULTS: Endoscopy was performed in 1983 patients of whom 219 (11.0%) had overt upper digestive tract hemorrhage. Reflux esophagitis was the cause of bleeding in 32 patients (14.6%). Reflux esophagitis was diagnosed in 391 patients during the same period of time. Bleeding reflux esophagitis accounted for 8.2% of them. Independent factors associated with bleeding were grade 3 or 4 (Savary-Miller) esophagitis (odds ratio [OR]: 25.5, 95% confidence interval [CI]: 9.6-67.9), cirrhosis (OR: 5.7, 95% CI: 1.7-18.9), Eastern Cooperative Oncology Group performance status > or = 3 (OR: 4.6, 95% CI: 1.5-14.2), and anticoagulant therapy (OR: 3.9, 95% CI: 1.2-12.5). A history of reflux esophagitis or heartburn was noted in only 28.1% or 37.5% of the patients with bleeding reflux esophagitis, respectively. CONCLUSIONS: In this population of patients with reflux esophagitis, the prevalence of bleeding esophagitis was high (8.2%). Bleeding esophagitis occurred primarily in patients with severe esophagitis and was the revealing clinical form of gastroesophageal reflux disease in the majority of cases, suggesting that bleeding prevention would hardly be effective.