Atrial fibrillation (AF) is the most common cardiac arrhythmia in the general population and in patients with a history of cardiovascular disease. AF is becoming an outbreak particularly for the western countries as it increases with advancing age; furthermore, AF has a negative social impact because it is associated with stroke and myocardial infarction. Thrombosis generated in the left atrial appendage with ensuing embolism in the cerebral circulation is considered the most important cause of ischemic stroke. In addition to thrombo-embolism, AF is characterized by a constellation of atherosclerotic risk factors, including hypertension, dyslipidaemia and diabetes, which may predispose to serious clinical complications of atherosclerosis such myocardial infarction. Even if interventional trials with oral anticoagulants such as warfarin reduced by about 60% the risk of stroke, AF patients still disclose an elevated residual cardiovascular risk, which may severely complicate the clinical course and management of AF. Recent trials with new oral anticoagulants (NOACs) are opening a new scenario for the treatment of AF, which could improve its management, as NOACs apparently would not require monitoring. However, important caveats are emerging in the real world of AF management, which are questioning the concept that NOACs do not need monitoring. Thus, issues related to compliance and large variability in blood concentration may negatively influence the cost/effectiveness benefit of NOACs. This review will focus on pathophysiology of thrombo-embolism and athero-thrombosis and the impact of old and new anticoagulants in the real world of AF management.