The aim of this study was to determine the relationship between alcohol consumption and occurrence of peripheral arterial disease in the general population. During 1988 in a cross sectional survey, the Edinburgh Artery Study, 1,592 men and women aged 55-74 years were selected at random from the age-sex registers of ten general practices distributed geographically and socio-economically across the city. participants were asked to recall the number of units of wine, beer and spirits consumed in the previous week and whether or not this was typical. Peripheral arterial disease was measured using the ankle brachial pressure index (ABPI). Men and women were analysed separately because of large differences in alcohol consumption. There was no association between ABPI and alcohol consumption in women but, in men, increasing alcohol consumption was associated with a higher ABPI (test for trend, p = 0.03) indicating less severe disease. This relationship was linear rather than U-shaped. In multiple regression analysis, after age-adjustment the ABPI was related to wine consumption but not beer or spirits in men (p < or = 0.01). On adjusting for age and cumulative lifetime cigarette smoking, the association of wine consumption with the ABPI was diminished but remained statistically significant (p < 0.05). On adjusting for age and social class, the relationship of total alcohol intake and wine consumption with the ABPI became non significant (p > 0.05). We conclude that in males, greater alcohol consumption is related to a higher ABPI and that any protective 'effect' of alcohol relates to wine consumption rather than beer or spirits.(ABSTRACT TRUNCATED AT 250 WORDS)