Background: Ximelagatran is a novel oral direct thrombin inhibitor that is as effective as warfarin in preventing stroke and other thromboembolic complications in patients with nonvalvular atrial fibrillation (AF). Risk factors for bleeding with warfarin are known, but risk factors for bleeding with ximelagatran have not been described. Unlike warfarin, ximelagatran has a predictable anticoagulant effect, does not require anticoagulation monitoring, has a low potential for interactions with drugs, food, or alcohol, and is not affected by genetic polymorphisms. We undertook an exploratory analysis of a large patient database to identify conventional and novel risk factors for bleeding in ximelagatran-treated patients, in warfarin-treated patients, and in all patients, irrespective of treatment.Methods: We undertook a pooled analysis of the SPORTIF III and V trials trials, which included 7329 patients with nonvalvular AF who received oral ximelagatran, 36 mg twice daily, or warfarin, administered to achieve a target international normalized ratio of 2.0–3.0. Patients had nonvalvular AF and 1 or more risk factors for stroke: hypertension; age ≥75 years; previous stroke, transient ischemic attack (TIA) or systemic embolism; left ventricular dysfunction; age ≥65 years and coronary artery disease; or age ≥65 years and diabetes mellitus. Major exclusion criteria were: mitral stenosis; previous heart valve surgery; transient AF; increased risk for bleeding. Multivariate logistic regression analysis was used to identify independent risk factors for major bleeding. The hazard ratio (HR) for major bleeding, and corresponding 95% confidence interval (CI), was calculated for each variable in the regression model.Results: The Table presents risk factors in which there was a significant or a non-significant (NS) association with major bleeding in ximelagatran-treated or warfarin-treated patients, and in the combined patient population.Risk factorXimelagatran-treated patients, HR (95% CI)Warfarin-treated patients, HR (95% CI)Combined patient population, HR (95% CI)Aspirin use1.65 (1.07, 2.55)2.40 (1.69, 3.42)1.96 (1.49, 2.58)Increasing age1.03 (1.01, 1.05)1.06 (1.03, 1.08)1.04 (1.03, 1.06)Prior liver diseaseNS4.96 (1.57, 15.62)NSPrior stroke or TIA1.78 (1.16, 2.73)NSNSDiabetes mellitus1.80 (1.18, 2.75)NS1.39 (1.05, 1.86)Asian raceNSNS1.99 (1.16, 3.42)Statin use0.62 (0.39, 0.97)0.61 (0.42, 0.88)0.62 (0.39, 0.97)Conclusions: Overall, the bleeding risk was lower with ximelagatran compared with warfarin. Aspirin use and increasing age were associated with an increased risk of bleeding in both ximelagatran- and warfarin-treated patients. Statin use was associated with a decreased risk for bleeding in both groups.