Between 1993 and 1997, 16 165 women aged 49–70 years and free from cardiovascular disease were enrolled in the Dutch Prospect-EPIC cohort (European Prospective Study Into Cancer and Nutrition) and followed up for a median period of 75 months. At enrolment, women filled in questionnaires on chronic disease risk factors and nutrition. Intake of phyto-oestrogens was estimated using the food frequency questionnaire covering regular dietary intake of 178 food items in the year before enrolment. Cox regression analysis was used to estimate hazard ratios of cardiovascular disease for quartiles of phyto-oestrogen intake adjusted for age at intake, body-mass index, smoking, physical activity, hypertension, hypercholesterolaemia, use of hormone replacement therapy, menopausal status and intake of total energy, total fibre, vegetables, fruit and alcohol. In total, 372 women experienced a coronary event and 147 women a cerebrovascular event during follow-up. Overall, neither isoflavones nor lignans were associated with decreased cardiovascular disease risk. When stratifying for ever vs. never smokers, the risk of coronary events decreased with increasing lignan intake for ever smokers.