An association between alcohol consumption and blood pressure levels has been observed in over 60 population studies world wide. The relationship is generally linear but with some studies showing a threshold effect at around 2-3 standard drinks a day. Effects are seen with all types of alcoholic beverages, and in men and women. The affect appears additive to effects of obesity and higher dose oral contraceptives. Studies of acute effects of alcohol suggest an initial vasodilator response, while population studies suggest that heavy drinkers may show some rebound hypertension. Randomized controlled trials show that reducing alcohol consumption lowers blood pressure in both treated and untreated hypertensives. Mechanisms of alcohol induced hypertension are still unclear. Despite predisposing to hypertension, regular light to moderate drinking (1-4 standard drinks a day) appears to protect against coronary deaths and ischaemic strokes, while heavier drinking increases the risk of haemorrhagic stroke and heart disease. There is some suggestion that wine drinking may be associated with lower cardiovascular risks, however in a study of 343 working men we found that beverage preference and drinking patterns correlated strongly with diet habits, smoking education and socioeconomic status, factors that are likely to confound the interpretation of epidemiological studies suggesting favourable cardiovascular effects to a particular beverage. Although light drinkers have lower mortality than non drinkers those drinking more than 2 standard alcohol drinks per day show a rising mortality as well as an increased risk of hypertension. Those facts should be the basis of public health advice on 'safe' levels of drinking.