Recent epidemiologic studies have consistently shown that moderate intake of alcoholic beverages protect against morbidity and mortality from coronary heart disease and ischemic stroke. By contrast, alcohol drinking may also predispose to cerebral hemorrhage. These observations suggest an effect of alcohol similar to that of aspirin. Several studies in humans and animals have shown that the immediate effect of alcohol, either added in vitro to platelets or 10 to 20 min after ingestion, is to decrease platelet aggregation in response to most agonists (thrombin, ADP, epinephrine, collagen). Several hours later, as, in free-living populations deprived of drinking since the previous day it is mostly secondary aggregation to ADP and epinephrine and aggregation to collagen that are still inhibited in alcohol drinkers. By contrast, in binge drinkers or in alcoholics after alcohol withdrawal, response to aggregation, especially that induced by thrombin, is markedly increased. This rebound phenomenon, easily reproduced in rats, may explain ischemic strokes or sudden death known to occur after episodes of drunkenness. The platelet rebound effect of alcohol drinking was not observed with moderate red wine consumption in man. The protection afforded by wine has been recently duplicated in rats by grape tannins added to alcohol. This protection was associated with a decrease in the level of conjugated dienes, the first step in lipid peroxidation. In other words, wine drinking does not seem to be associated with the increased peroxidation usually observed with spirit drinking. Although further studies are required, the platelet rebound effect of alcohol drinking could be associated with an excess of lipid peroxides known to increase platelet reactivity, especially to thrombin.