Atrial fibrillation is the most common clinically significant arrhythmia observed both in the general population and in competitive athletes. The most important risk factors are all preventable by regular physical activity. However, although the benefits of moderate physical activity in controlling cardiovascular risk factors and decreasing the risk of atrial fibrillation have been extensively proved, concerns have arisen about the potential negative effects of vigorous exercise, particularly in endurance athletes. Furthermore, in a subset of patients with atrial fibrillation younger than 60 years, routine evaluation does not reveal any cardiovascular disease or any other known causal factor. This condition is called 'lone atrial fibrillation', and the potential mechanisms underlying this condition are speculative and remain to be clarified. Atrial ectopy, increased vagal tone, changes in electrolytes, left atrial dilatation, and fibrosis have been proposed among others as potential mechanisms. However, no convincing data still exist. Particularly, the increase in left atrial size represents in athletes a physiological adaptation to exercise conditioning and the presence of biatrial fibrosis has not been demonstrated in humans. Thus, contrary to patients with cardiovascular disorders, the atrial substrate seems to play a secondary role in healthy athletes. This review article analyzes the controversial relationship between atrial fibrillation and physical activity, with a particular attention on the pathophysiological mechanisms that could be responsible for atrial fibrillation in the athletic population.