BACKGROUND: The optimal treatment of ulcers with nonbleeding visible vessels and adherent clots is unclear. OBJECTIVE: To compare intravenous omeprazole infusion plus endoscopic therapy with intravenous omeprazole infusion alone for prevention of recurrent bleeding from ulcers with nonbleeding visible vessels or adherent clots. DESIGN: Single-blind randomized study with blinded evaluation of study end points. SETTING: An endoscopy center in a university hospital in Hong Kong. PATIENTS: 156 persons with upper gastrointestinal bleeding and ulcers showing nonbleeding visible vessels or adherent clots. INTERVENTION: Combination of endoscopic therapy and omeprazole infusion versus sham endoscopic therapy and omeprazole infusion. MEASUREMENTS: Recurrent ulcer bleeding before discharge and within 30 days. RESULTS: 78 patients were recruited in each group. Ulcer bleeding recurred before discharge in seven patients who received intravenous omeprazole alone (9%) and no patients who received combined therapy (difference, 9 percentage points [95% CI, 1.7 to 17.6 percentage points]; P = 0.01). The probability of recurrent bleeding within 30 days was 11.6% (9 patients) in the omeprazole-alone group and 1.1% (1 patient) in the combined therapy group (difference, 10.5 percentage points [CI, 1.7 to 19.8 percentage points]; P = 0.009). Patients in the combined therapy group required less transfusion (difference in median units of blood transfused, 1 unit [CI, 0 to 2 units]; P = 0.02). One patient in the combined therapy group had surgery for ulcer perforation. Four patients receiving omeprazole alone (5.1%) and two patients receiving combined therapy (2.6%) died within 30 days. CONCLUSION: The combination of endoscopic therapy and omeprazole infusion is superior to omeprazole infusion alone for preventing recurrent bleeding from ulcers with nonbleeding visible vessels and adherent clots.