This thesis has explored the information content of the ventricular fibrillation (VF) waveforms recorded as either a surface electrocardiogram (ECG) or an intracardiac electrogram (EGM). Ventricular fibrillation is lethal if not promptly terminated resulting in very low survival rates especially if the arrest occurs outside of a hospital setting. The current treatment strategy is defibrillation. However, even if performed in a timely manner, this is not always successful. It is hoped that further insight into the appearance of VF waveforms and what, if any, clinical information these hold may improve survival rates by allowing additional treatment strategies to be employed. A key step in this process is interpreting the information that can be extracted from VF waveform analysis and the relationship that this has with underlying cardiac pathology and clinical outcomes. Chapter 1: Provides a background context for the work presented. Chapter 2: Examined the variation in VF waveform characteristics across a standard 12 lead ECG. We demonstrated that while there are similarities between each of the surface leads additional information was added by analysing multiple recordings. Chapter 3: Compared the information content of several commonly used analysis tools, from which we have created our own set of measures. These extract information regarding the rate, variance in the rate and variance in the waveform shape. Chapter 4: The analysis tools determined in Chapter 3 were used to demonstrate a reduction in VF frequency in the presence of ischemic scar, decreased left ventricular ejection fraction and a dilated left ventricle. Chapter 5: The same analysis tools were then used to demonstrate that failed ICD shocks were preceded by episodes of faster VF with less regular waveform appearance than those that were successfully terminated. Chapter 6: We investigated the potential relationship between VF waveform characteristics at the time of ICD implant and the occurrence of subsequent significant ventricular arrhythmia. We were unable to demonstrate a link between the two using appropriate shock treatment to estimate the ventricular arrhythmia burden. Chapter 7: A unique analysis tool was investigated to describe VF waveform regularity. Standard 12 lead ECGs were used to derive a 3 dimensional plot of the excitation wave front for each VF cycle. Large variation in the waveform regularity was observed. Chapter 8: A summary of the key findings learnt in Chapters 2 to 7. The clinical implications of this work and possible future directions are discussed. Chapter 9: References quoted in this thesis.