Studies in experimental animals have shown varying and apparently opposite effects of alcohol on cardiac rhythm and conduction. Given acutely to non-alcoholic animals, ethanol may even have anti-arrhythmic properties whereas chronic administration clearly increases the animals' susceptibility to cardiac arrhythmias. Chronic heavy alcohol use has been incriminated in the genesis of cardiac arrhythmias in humans. The evidence has come from clinical observations, retrospective case-control studies, controlled studies of consecutive admissions for arrhythmias, and prospective epidemiological investigations. Furthermore, electrophysiological studies have shown that acute alcohol administration facilitates the induction of tachyarrhythmias in selected heavy drinkers. The role of alcohol appears particularly conspicuous in idiopathic atrial fibrillation. Occasionally, ventricular tachyarrhythmias have also been provoked by alcohol intake. Several lines of evidence suggest that heavy drinking increases the risk of sudden cardiac death with fatal arrhythmia as the most likely mechanism. According to epidemiological studies this effect appears most prominent in middle-aged men and is only partly explained by confounding traits such as smoking and social class. The basic arrhythmogenic effects of alcohol are still insufficiently delineated. Subclinical heart muscle injury from chronic heavy use may be instrumental in producing patchy delays in conduction. The hyperadrenergic state of drinking and withdrawal may also contribute, as may electrolyte abnormalities, impaired vagal heart rate control, repolarization abnormalities with prolonged QT intervals and worsening of myocardial ischaemia or sleep apnoea. Most of what we know about alcohol and arrhythmias relates to heavy drinking. The effect of social drinking on clinical arrhythmias in non-alcoholic cardiac patients needs to be addressed further.