Light-to-moderate alcohol intake is associated with a reduced incidence of ischaemic cardiovascular events, whilst heavy alcohol intake can predispose individuals to stroke. Alcohol-induced changes in coagulation and fibrinolysis may be relevant and are the subject of this controlled trial of varying alcohol intake in 55 predominantly beer-drinking men. Following 4 weeks stabilization maintaining usual drinking habits, participants were randomized to either continue usual alcohol intake or to restrict alcohol by changing to low alcohol beer for 4 weeks. In a final 4 week period, they crossed over to low or usual alcohol intake, respectively. Comparing combined low and usual alcohol periods, an increase in mean weekly alcohol intake from 92 to 410 ml (mean daily intake from 13 to 58 ml) was associated with a decrease in plasma fibrinogen (by 11%, P < 0.001) and platelet count (3%, P < 0.05), but increases in factor VII (7%, P = 0.001), tissue plasminogen activator (tPA; 16%, P = 0.01) and plasminogen activator inhibitor-1 (PAI-1; 21%, P < 0.001). The ratio, tPA/PAI-1, fell from 0.50 to 0.44 (P = 0.02) confirming the relatively greater increase in PAI-1 with alcohol consumption. Two lipid-associated natural anticoagulants, tissue factor pathway inhibitor and beta 2-glycoprotein-I, did not change. The substantial reduction in plasma fibrinogen with alcohol intake may well contribute to the apparent protection alcohol confers against ischaemic coronary and cerebral events. The increase in factor VII and relatively greater increase in PAI-1 than tPA with alcohol intake may attenuate this benefit and indeed may sufficiently predispose individuals to thrombosis to contribute to the increased incidence of ischaemic stroke seen in heavier drinkers. The balance of anticoagulant and procoagulant and fibrinolytic effects in any individual may vary depending on quantity and type of alcoholic beverage ingested, as well as on genetic and other variables, all of which merit further study.