Atrial fibrillation (AF) is a rapid and irregular activation of the atria that represents by far the most common cardiac arrhythmia. Despite the filtering effect of the atrioventricular node, transmission of excitation to the ventricles results in an inappropriately high ventricular rate during AF while the pooling of immotile blood within the atrial chambers leads to an increased risk of thrombo-embolic events. In consequence, AF contributes to the development of heart disease, is a major cause of stroke and is thereby associated with increased mortality. AF is not only a factor in the development of heart disease but can also be caused by the existence of a predisposing cardiac condition: for example, the presence of systemic hypertension and left ventricular hypertrophy are good epidemiological indicators of the risk of developing AF. Valve disease and regurgitation are also risk factors for the development of AF. The association of AF with pre-existing structural heart disease has led to the suggestion that remodelling of the atria in heart disease, possibly involving increased haemodynamic load on the atrial wall, predisposes the atrium to AF. The mechanisms underlying this remodelling remain unclear and it is hoped that an improved understanding of the processes leading to the genesis of AF might lead to new therapeutic strategies in the management of at-risk patients. However, the aetiology of the disease is usually complex and it is extremely difficult to study the mechanisms underlying the pathogenesis of AF in patients. On the other hand, models of heart disease in small animals (e.g. rats) can afford an accessible and cost-effective means of studying in the laboratory the mechanisms underlying pro-arrhythmic atrial remodelling caused by a single disease process. Multiple electrode array recording of atrial electrograms from the epicardial surface of isolated perfused hearts allows the measurement of atrial effective refractory period and conduction velocity and the assessment of localised inhomogeneities in atrial conduction. The inducibility of AF can be assessed by recording the incidence of paroxysms of arrhythmia following burst pacing of the atrium. All animal procedures were considered by the research ethics committee of the University of Bristol and were in accordance with the Animals (Scientific Procedures) Act, 1986 of the United Kingdom. Data from studies of atrial remodelling in an in-bred model of systemic hypertension, the spontaneously hypertensive rat (SHR), and in a surgical model of elevated afterload involving banding of the ascending aorta in rats, in comparison with the appropriate controls demonstrate the importance of fibrosis and conduction abnormalities in forming a substrate for re-entrant tachyarrhythmia and increased inducibility of AF (Choisy et al., 2007; Kim et al., 2011). In both models, the pro-arrhythmic atrial remodelling involved the long term elevation of afterload, being evident in hearts from 11 month old SHR but not at 3 months of age and being evident after 20, but not at 8, weeks post surgery in the aortic banding model. Notably, while treatment of SHR with the vasodilator, hydralazine (14 mg/kg/day), or the angiotensin receptor blocker, candesartan (3 mg/kg/day), via the drinking water for 14 weeks prior to experimentation led to the normalisation of arterial pressures and the regression of atrial and ventricular hypertrophy, fibrosis remained unchanged, suggesting that it may be difficult to reverse pro-arrhythmic structural remodelling. It is concluded that small animal models of heart disease represent a useful tool for the investigation of the mechanisms underlying pro-arrhythmic atrial remodelling in heart disease.