OBJECTIVE: To determine whether the rate of cardiovascular disease is different among parous women with a general practitioner reported history of toxaemia of pregnancy than among those not reported to have experienced toxaemia, or among nulliparous women. DESIGN: Prospective cohort study. SETTING: 1400 general practitioners throughout the United Kingdom. SUBJECTS: Women who had never used oral contraceptives who were recruited to the Royal College of General Practitioners' oral contraception study (original cohort about 23000). MAIN OUTCOME MEASURES: Age, social class, and smoking standardised incidence rates for hypertensive disease, acute myocardial infarction, other acute ischaemic heart disease, other chronic ischaemic heart disease, angina pectoris, total ischaemic heart disease, total cerebrovascular disease, and total venous thromboembolic disease in the three groups. RESULTS: Compared with parous women with no history of toxaemia, those who had experienced toxaemia had a significantly increased risk of hypertensive disease (relative risk (RR) 2.35), acute myocardial infarction (RR 2.24), chronic ischaemic heart disease (RR 1.74), angina pectoris (RR 1.53), all ischaemic heart disease (RR 1.65), and venous thromboembolism (RR 1.62). The rates for all cerebrovascular disease and peripheral vascular disease were also increased but not significantly. Nulliparous women were more likely to develop hypertension or all cerebrovascular disease later in life than parous women without a history of toxaemia. CONCLUSIONS: A history of toxaemia of pregnancy increases the risk of several distinct cardiovascular conditions later in life. Although causality cannot be inferred (other characteristics of the women may account for both an increased risk of toxaemia and a risk of subsequent vascular disease), the findings merit further research because of their potential importance.