The relationships between alcohol intake and blood pressure have been examined in 7,735 middle-aged men drawn at random from general practices in 24 British towns. Both mean systolic and diastolic BP are increased in moderate (16-42 drinks/week) and heavy (greater than 42 drinks/week) drinkers. The alcohol-blood pressure relationship is independent of age, body mass index and social class despite their associations with both alcohol intake and BP. The prevalence of hypertension (systolic greater than or equal to 160 mmHg or diastolic greater than or equal to 90 mmHg) is increased in both moderate and heavy drinkers. Non-drinkers have higher rates of diastolic hypertension than occasional or light drinkers, probably due to changes in drinking habits made by those diagnosed as hypertensives. Furthermore, recall of doctor diagnosis of hypertension and of anti-hypertensive treatment is highest among non-drinkers. It is estimated that about 10% of hypertension (systolic or diastolic) can be attributed to moderate or heavy drinking. There is a clear need for increased awareness of the alcohol-blood pressure relationship. Current drinking status should be determined in all hypertensive subjects, and assessment of alcohol effect by temporary withdrawal should be the first step in the management of anyone with sustained hypertension who drinks.