Background: Patients with atrial fibrillation (AF) who are new to anticoagulation (warfarin-naive [WN]) may have a higher risk of thrombosis and/or bleeding than patients who are warfarin-experienced (WE). In the ARISTOTLE trial, centers were asked to include at least 40% WN patients. We assessed the treatment effect of apixaban versus warfarin in WN and WE cohorts. We also tested the hypothesis that, overall, the risk of adverse outcomes would be higher in the WN cohort. Methods: Patients were defined as WN if they had used warfarin for Results: Of all 18,201 pts enrolled, 7,800 (43%) were WN and 10,401 were WE. Compared to WE pts, WN pts were more often female (39% vs. 33%). WN and WE pts were similar with respect to age and CHADS 2 score. At baseline, fewer WN pts had a history of or prior stroke (18% vs. 21%) and prior bleeding (10% vs. 22%). During the 1.8 years of randomized treatment compared to warfarin, apixaban reduced the risk of stroke/SE and caused less bleeding, irrespective of WN status (table). In an analysis using data from both randomized treatment arms, WN pts had a numerically higher rate of stroke and systemic embolism (adjusted HR = 1.23; 95%CI: 1.00 - 1.52; p=0.06) and a similar rate of major bleeding in comparison to WE patients. Premature study drug discontinuation was more common in WN patients (1.19; 95%CI: 1.10-1.28; p Conclusions: Compared to warfarin, apixaban reduced the rates of stroke/SE and major bleeding in both warfarin-naive and warfarin-experienced patients with AF. Overall, we observed a trend toward a higher rate of stroke/SE in warfarin-naive (compared to warfarin-experienced) pts.