OBJECTIVES: The American College of Chest Physicians, the European Society of Cardiology guidelines, and published decision support tools by LaHaye and Casciano offer recommendations to guide oral anticoagulant (OAC) treatment in patients with atrial fibrillation (AF).  The aim of our study is to compare the predictive validity of these aids by contrasting the net clinical benefit when OAC use is concordant/discordant with each of the aids.  METHODS: A cohort study using the 2006-2013 Lifelink claims data was used to compare the net clinical benefit (NCB) of AF patients. NCB is the difference in event rates of composite events (thromboembolic stroke and major bleed events per 10,000 person years) between patients who are concordant and those who are discordant with the guideline/tool recommendations. Cox proportional hazard models were used to assess the relative risk of composite thromboembolic and bleed events adjusted for potential confounders. RESULTS:  A total of 11,315 AF patients contributing 27,801 person years met the study inclusion criteria. The NCB of patients concordant with recommendations of the LaHaye tool (34.81[C.I=33.4-36.23]) was highest followed by American guidelines (22.75[C.I=21.40-24.17]), Casciano tool (16.99[C.I=15.57-18.40]) and European guidelines (3.94[C.I=2.52-5.35]). By restricting the definition of composite events to ischemic stroke and intracranial hemorrhage; the NCB of patients concordant with American guidelines (24.20[C.I=22.88-25.52]) was highest followed by Casciano tool (17.07[C.I=15.75-18.38]), LaHaye tool (14.30[C.I=12.99-15.62]) and European guidelines (11.25[C.I=9.93-12.57]). There was no significant decrease in the risk of composite events associated with concordant OAC use/non-use for any of the decision aids after multivariate adjustment.    CONCLUSIONS:  These results suggest that OAC use/non-use consistent with any of the tools led to net clinical benefits but the rank order depended on the composite outcomes selected.  However, the benefit vanished after multivariate adjustment.  Larger studies are warranted before any one OAC decision aid can be recommended to routinely guide OAC treatment decisions.