Disturbed autonomic nervous 'balance' of the sympathetic nervous and vagal outflows to the heart potentiates the experimental development of ventricular arrhythmias in laboratory animals. For some time the best evidence for the occurrence of a similar phenomenon in humans was provided by the long QT interval syndrome, sufferers of which are very prone to develop serious ventricular arrhythmias and in whom evidence exists of abnormal anatomy and function of the cardiac sympathetic nerves. Recently the case for disturbed autonomic function causing clinical arrhythmias has become more broadly based. Reduced baroreflex sensitivity after myocardial infarction, and low heart rate variability, both of which rest largely on vagal underactivity, have been shown to be associated with substantially increased risk of subsequent sudden death. A second observation is that patients having recovered from unexpected ventricular tachycardia or ventricular fibrillation have markedly increased cardiac sympathetic activity compared with appropriate reference groups, based on measurements of the rate of spillover of the sympathetic neurotransmitter, noradrenaline, from the heart to plasma. These clinical findings support a role for cardiac autonomic dysfunction, specifically sympathetic activation and vagal withdrawal, in arrhythmogenesis. These observations are timely, given the recent demonstration that most conventional anti-arrythmics are of little benefit in preventing sudden death. A reappraisal of the anti-arrhythmic activity of beta-adrenergic blocking drugs, and evaluation of potential benefits of other pharmacological and non-pharmacological means of favourably altering cardiac autonomic function is now needed.