PURPOSE: To evaluate the accuracy and clinical utility of the Outpatient Bleeding Risk Index for estimating the probability of major bleeding in outpatients treated with warfarin. The index was previously derived in a retrospective cohort of 556 patients from a different hospital (derivation cohort). SUBJECTS AND METHODS: We enrolled 264 outpatients starting warfarin (validation cohort) to validate the index prospectively. All patients were identified upon hospital discharge, and physician estimates of the probability of major bleeding were obtained before discharge in the validation cohort. RESULTS: Major bleeding occurred in 87 of 820 outpatients 0 (6.5%/yr). The index included four independent risk factors for major bleeding: age 65 years or greater; history of gastrointestinal bleeding; history of stroke; and one or more of four specific comorbid conditions. In the validation cohort, the index predicted major bleeding: the cumulative incidence at 48 months was 3% in 80 low-risk patients, 12% in 166 intermediate-risk patients, and 53% in 18 high-risk patients (c index, 0.78). The index performed better than physicians, who estimated the probability of major bleeding no better than expected by chance. Of the 18 episodes of major bleeding that occurred in high-risk patients, 17 were potentially preventable. CONCLUSIONS: The Outpatient Bleeding Risk Index prospectively classified patients according to risk of major bleeding and performed better than physicians. Major bleeding may be preventable in many high-risk patients by avoidance of over-anticoagulation and nonsteroidal anti-inflammatory agents.