Atrial fibrillation (AF) is common and is associated with significantly increased morbidity and mortality (mostly due to heart failure and stroke). A striking feature of AF is the ability of this arrhythmia to sustain itself. It is accepted that, by increasing atrial rate, AF leads to electrical and structural remodelling of the atria, which, in turn, promotes AF maintenance and increases vulnerability to relapse. There is considerable interest in developing treatment and prevention strategies that target mechanisms upstream of ion channels modifications (e.g., myocardial redox state and metabolism); however, whereas changes in ion channels conductance are the common denominator of virtually all types of AF, the myocardial signalling upstream of atrial electrical and structural remodelling may differ with the stage and substrate of AF, demanding a more refined ad hoc approach to the prevention and management of this arrhythmia. For example, we have shown that a gp91phox-containing NADPH oxidase (NOX2) is the main source of reactive oxygen species (ROS) in human atrial myocytes, and that atrial NOX2 activity is independently associated with an increased risk of developing AF and other postoperative complications in patients undergoing cardiac surgery. We have also established that short term treatment (3 days) or ex vivo tissue incubation with HMG CoA reductase inhibitors (statins) increases NO bioavailability and reduces vascular and cardiac NOX2 activity in patients undergoing coronary revascularization by inhibiting one of the key components of the oxidase complex (i.e., the small G protein Rac1), before any change in LDL cholesterol can be detected. Together, these findings suggest that NOX2 inhibition with statins may help prevent the new onset of AF after cardiac surgery and, possibly, in other disease states characterized by inflammation. However, whether NOX2 inhibition may also be beneficial in the secondary prevention of AF or in conditions where atrial structural remodeling provides the arrhythmic substrate remains a matter of debate. To address this issue, we investigated the time-dependent changes in the expression and activity of atrial oxidase systems in the atrial myocardium of goats with pacing-induced AF (2 weeks and 6 months) or atrial structural remodeling secondary to atrioventricular block, and in human atrial samples. We found that NOX2 and Rac1 activity and the protein level of the cytochrome-forming subunits of the oxidase (NOX2 and p22phox) are significantly increased in the left atrial myocardium of goats after 2 weeks of AF and in atrial samples from patients in sinus rhythm who develop AF after cardiac surgery. By contrast, the increase in ROS production observed in longstanding AF or in the presence of atrio-ventricular block (i.e., in the presence of atrial structural remodeling) is due to mitochondrial oxidases and NOS “uncoupling” (a phenomenon whereby the catalytic electron flow within the enzyme is uncoupled from NO synthesis and diverted to molecular oxygen to yield superoxide), secondary to a reduction in the atrial content of the NOS co-factor BH4 and an increase in arginase activity. Ex vivo atorvastatin causes a mevalonate-reversible inhibition of atrial Rac1 and NOX2 activity in patients who develop AF after cardiac surgery, but it does not affect atrial ROS production, NOS uncoupling or BH4 content in patients with permanent AF. Taken together, these data demonstrate that the mechanisms responsible for the NO-redox imbalance in the fibrillating atrial myocardium change with the duration of AF and the development of atrial structural remodeling and suggest that a shift in the atrial sources of ROS with the duration of AF may influence the antiarrhythmic efficacy of statins.