Alcohol intake has been associated with higher blood pressure in acculturated populations but not in unacculturated societies. We performed a cross-sectional survey of a random community sample of 5023 male Yi rural farmers and 1656 Yi and 2173 Han men living in an urban setting. Average alcohol intake among drinkers was 36.4 g/d in Yi farmers, 56.5 g/d in Yi migrants, and 38.7 g/d in Han men. Age-adjusted mean diastolic blood pressure was 66.9, 70.5, and 71.7 mm Hg, respectively. Diastolic blood pressure was higher at higher alcohol intakes in all three groups (all P < .001). After adjustment for age, body mass index, heart rate, smoking, and physical activity, the change (95% confidence interval) in diastolic blood pressure for each standard drink was 0.50 (0.38-0.62), 0.31 (0.18-0.43), and 0.24 (0.07-0.40) mm Hg for Yi farmers, Yi migrants, and Han men, respectively. The percentage of variance in diastolic blood pressure explained by alcohol intake was 5% for Yi farmers, 4% for Yi migrants, and 2% for Han men. In a random sample of 831 men, these associations were independent of urinary sodium, potassium, calcium, and magnesium and sodium-potassium ratio. In the Yi farmers, associations were less strong for systolic blood pressure and no longer significant after adjustment. Approximately 33% of hypertension could be attributed to daily alcohol use in the Yi groups compared with 9.5% in the Han people.(ABSTRACT TRUNCATED AT 250 WORDS)