The objective of this cross-sectional study was to analyse the influence of biological, socio-demographic, and psychosocial factors and current perimenopausal status on hypertension in a geographically defined population of 10,766 women aged 50-59 years, of whom 6901 attended the study. Altogether 1887 (27.3%) women had hypertension: 996 with drug treatment and 891 diagnosed at the study. In a logistic multiple regression analysis (controlled for age), drug treatment of hyperlipidaemia, family history of hypertension, waist-to-hip ratio, body mass index (BMI) increase > or = 25% during the past 25 years, S-triglycerides, S-cholesterol, education up to comprehensive school, and to upper secondary school, consumption of 84-167 g of alcohol/week, and of > or = 168 g of alcohol/week, were positively associated with hypertension, while high-density lipoprotein cholesterol and current smoking were negatively associated. A significant interaction was found between current smoking and BMI increase, with a lower risk for hypertension among smokers who had increased their BMI > or = 25%. No interaction was found between smoking and alcohol. In conclusion, hypertension was predominantly associated with biological factors, and with heredity for hypertension. Of the socio-demographic factors, only low level of education was associated with hypertension in a comprehensive analysis. Perimenopausal status showed no relation to occurrence of hypertension in the multiple regression analysis. The risk for hypertension increased with moderate and high consumption of alcohol, whereas smoking showed a decreased risk. Among women with weight gain, present smoking remained protective. Although both smoking and hypertension are established risk factors for cardiovascular disease, they seem not to be directly linked, indicating a complexity of mechanisms.