The risk of thromboembolic events, particu larly stroke, is increased in patients with atrial fibrillation. Therefore, it is vitally important to identify effective strategies for adequate antico agulation, especially for those at greatest risk. Determining which patient has a substantial risk and is most likely to derive benefit from anticoagulation without incurring excessive bleeding risk, however, is far from simple. The exact mechanisms and patient characteristics lending to this increased risk of thromboembo lism in atrial fibrillation are still being defined. Furthermore, current anticoagulation strategies are often cumbersome in clinical practice, replete with risks and restrictions, do not fully negate the risk of thrombo embolism and, as such, are unpalatable to patients and physicians alike. Several newer anticoagulants, including fac tor Xa antagonists and direct thrombin inhibi tors, appear promising despite the underlying complexities of atrial fibrillation. Recent data concerning dabigatran [1], a direct thrombin inhibitor, are compelling and give us pause to consider whether we are approaching a new fron tier in the management of thrombo embolic risk in atrial fibrillation patients. There are several described mechanisms responsible for the increased risk of thrombo embolic events in patients with atrial fibrilla tion [2]. One simple explanation is that inadequate atrial emptying, due to the rhythm itself, can cause stagnation of blood and a greater propen sity for thromboembolic events. Similarly, atrial remodeling as a result of fibrillation may increase the risk. Still, another consideration is that atrial fibrillation may lead to the activation of proco agulant pathways, perhaps based on an inflam matory response. Thromboemboli may also origi nate from noncardiac sources [3]. Despite these and other possible explanations, ongoing research seeks to better understand and characterize the responsible mechanisms so that treatment options may be developed to target these pathways. Comorbidities increase the risk of thrombo embolism in atrial fibrillation patients. Various scores that consider specific nonmodifiable clini cal characteristics have been developed to identify patients with nonvalvular disease at significant risk – the greater number of risk factors, the greater risk of stroke and need for anticoagulation. The CHADS 2 score, which assigns one point to the presence of congestive heart failure, hyper tension, age older than 75 years or diabetes, and two points to a history of stroke or transient ischemic attack, is commonly used to stratify the risk of thromboembolism in patients with nonvalvular atrial fibrillation. A score of 2 places a patient at an annual stroke risk of 4.0%; the annual risk of stroke is as high as 18.2% for those with a score of 6 [4–6]. While the annual risk of bleeding also varies by CHADS 2 score [7], ben efit in favor of warfarin anticoagulation increases when the score exceeds 1. The score, however, underestimates the substantial risk of thrombo embolism in some patients with atrial fibrillation. As a result, a more sophisticated and inclusive score, CHA 2 DS 2 VASc, incorporates female gen der, age of 65–74 years and presence of vascular disease as additional risk factors, each with its own impact [8]. Unfortunately, little attention is directed towards the paroxysmal and intermittent nature of atrial fibrillation in these risk stratification schemes. Patients may not be considered for anti coagulation after a documented return to sinus rhythm even though recurrent, but asymptom atic and occult atrial fibrillation puts patients at continued risk. In reality, any episode of atrial fibrillation has associated thrombo embolic risk, although the risk may be acceptably low. The use of an anticoagulant, like warfarin, involves evalu ating this risk and balancing it against the risk of bleeding [9], adverse effects and complexities of administration. The aforementioned risk stratification scores aid in determining which patients with atrial