Microvolt T-wave alternans is thought to predict the risk of ventricular arrhythmias in patients with heart disease, although recent clinical studies have conflicting results. Understanding the cellular basis for alternans may not only inform more effective utilisation of the clinical test, but also provide new insights into the causes of lethal arrhythmias in man. Cellular repolarisation alternans is thought to underlie T-wave alternans and in recent years, the concept of discordant repolarisation alternans has emerged as a new paradigm for the induction of re-entrant ventricular arrhythmia. This experimental observation has not been examined in clinically relevant models of pathology and so the aim of this study was to investigate whether increased transmural heterogeneity of repolarisation as a result of heart failure following myocardial infarction in the rabbit would predispose to the development of arrhythmogenic discordant alternans. A rabbit ventricular wedge preparation was developed and the transmural electrophysiology of intact rabbit ventricle was characterised using optical imaging techniques. This revealed transmural gradients of repolarisation in intact rabbit myocardium, which appeared to be influenced by electrotonic load, rather than purely being a reflection of intrinsic cellular differences. Interestingly, repolarisation alternans also appeared in transmural patterns, which were also modified by activation sequence, underlining the role of conduction and electrotonic influences in dictating the spatial patterns of alternans, which may be crucial in determining spatially discordant alternans. In this study, similar baseline electrophysiological characteristics were apparent in the remodelled myocardium of failing hearts compared with normal hearts, underlining the possible importance of dynamic factors in producing the increased vulnerability to re-entrant arrhythmias observed in failing hearts. Repolarisation alternans, elicited by low temperature and rapid pacing, occurred at lower heart rates in failing hearts. At physiological temperature, repolarisation alternans was also more common in failing hearts. Spatially discordant alternans was not consistently observed on the transmural surface and did not appear to be directly related to the development of arrhythmia. Failing hearts displayed an increased vulnerability to ventricular arrhythmia. Although heart failure was associated with both alternans and ventricular arrhythmia, there was no demonstrable mechanistic link between alternans and ventricular arrhythmias in failing hearts. These data establish the occurrence of repolarisation alternans in a clinically relevant pathology, and so constitute an important step forward in our understanding of the experimental paradigm. However, a definitive mechanistic link between alternans and arrhythmia in heart failure is yet to be shown.