OBJECTIVE To determine the relationship between obesity and subsequent incidence of ischaemic heart disease (IHD). METHODS Prospective cohort survey. METHODS Study of three occupational groups, with follow up examinations. METHODS 3500 people recruited between 1972 and 1978 (80% response rate), and followed up between 1978 and 1984. This report is based on subgroup of 1511 white men aged 40-64 at entry. RESULTS Information was obtained on smoking and family history of IHD. Blood pressure, weight, height, skinfold thickness at four sites, fibrinogen, factor VII activity and cholesterol were measured during follow up. Body mass index (BMI) was used as an index of obesity. BMI was found to be more strongly correlated with IHD than any of the skinfold measurements, none of which was significantly associated with IHD when BMI was allowed for. Increase in BMI by 1 SD (approximately 8 kg) was associated with a 44% increase in the risk of IHD. Of the four skinfolds, subscapular was the most closely associated with risk, confirming the relevance of central obesity. The association between obesity and IHD remained when possible mechanisms for its effects were taken into account, and its strength may increase with time: for 1 SD increase in BMI, risk of events within 5 years was increased by 28%, while risk of events after longer than 5 years was increased by 65%. CONCLUSIONS Preventive strategies for IHD should include avoidance of obesity.