BACKGROUND: Alcohol has been related to atrial fibrillation (holiday heart syndrome), but its electrophysiologic actions remain unclear. METHODS: We evaluated the effects of alcohol in 23 anesthetized dogs at baseline and after 2 cumulative intravenous doses of ethanol: first dose 1.5 ml/kg (plasma level 200 mg/dl); second dose 1.0 ml/kg (279 mg/dl). In 13 closed-chest dogs (5 with intact autonomic nervous system, 5 under combined autonomic blockade and 3 sham controls), electrophysiologic evaluation and monophasic action potential (MAP) recordings were undertaken in the right atrium and ventricle. In 5 additional dogs, open-chest biatrial epicardial mapping with 8 bipoles on Bachmann's bundle was undertaken. In the remaining 5 dogs, 2D echocardiograms and ultrastructural analysis were performed. RESULTS: In closed-chest dogs with intact autonomic nervous system, ethanol had no effects on surface electrocardiogram and intracardiac variables. At a cycle length of 300 milliseconds, no effects were noted on atrial and ventricular refractoriness and on the right atrial MAP. These results were not altered by autonomic blockade. No changes occurred in sham controls. In open-chest dogs, ethanol did not affect inter-atrial conduction time, conduction velocity, and wavelength. Atrial arrhythmias were not induced in any dog, either at baseline or after ethanol. Histological and ultrastructural findings were normal but left ventricular (LV) ejection fraction decreased in treated dogs (77 vs. 73 vs. 66%; p = 0.04). CONCLUSION: Ethanol at medium and high doses depresses LV systolic function but has no effects on atrial electrophysiological parameters. These findings suggest that acute alcoholic intoxication does not directly promote atrial arrhythmias.