The management and prevention of Sudden Cardiac Death remains a great challenge in modern Cardiology. Implantable Cardioverter Defibrillators (ICDs) have been shown to reduce mortality. Despite decades of research, the mechanisms are not fully understood and ICD treatment is crude, palliative and expensive. Nonetheless, outcome studies have helped to inform national and international guidance in the implantation of these devices. Patient selection is crucial to ensure correct patients are identified and appropriately treated. More refined and stringent risk stratification is needed to identify patients at high risk. This thesis examines non-invasive, readily measureable markers to see whether they can be used to assess the risk of ventricular arrhythmia in patients with cardiomyopathy who have indications for ICD implantation. Baseline data in the form of 12 lead electrocardiograms, echocardiography, 24 hour Holter monitoring and venous blood were obtained to analyse QT dispersion, Heart Rate Variability (HRV), QT Variability Index (QTVI), ECG restitution measures and NTproBNP levels in these patients. Patients were followed up for a two year period through the ICD clinic and appropriate therapy was recorded as a surrogate marker for ventricular arrhythmia. Patients with and without appropriate therapy were then compared to look for significant differences in the examined markers. The percentage of beats with a QT/TQ ratio>1 was associated with appropriate shocks when compared with no therapy (p=0.04). However, the result was not significant when all appropriate ICD therapy was compared with no therapy (p=0.06). This possibly reflects the period of time the heart spends on the more ‘unstable portion’ of the restitution slope in patients at highest risk. Median BNP was non-significantly higher in patients with arrhythmia compared to those who were shock free. None of the other examined markers were predictive of appropriate therapy. There is thus promise in the use of some non-invasive markers in the refinement of patient selection with LVSD being considered for ICD therapy.