The authors report the results of a collaborative study of the prevalence of ischaemic heart disease in 6 samples of middle-aged male clerical workers, drawn from 5 European countries and covering in all 4522 subjects. Detailed standardization of data-collection methods and of coding, analysis and tabulation was undertaken.Estimated prevalence rates for angina and history of possible infarction, based on the use of a standard questionnaire, showed significant variation between samples, but the validity of international comparisons of symptoms remains unproven. Prevalence rates for ischaemic-type ECG findings were high everywhere, and mostly showed no significant variation between the groups studied. Central coding of ECG findings proved essential to the making of unbiased comparisons. Neither the prevalence rates for symptoms nor those for ECG findings in these samples were recognizably correlated with the corresponding national mortality rates for ischaemic heart disease.Several possible risk factors were studied, but the only one found to be consistently associated with ischaemic heart disease was a raised blood pressure. Cigarette smoking was strongly associated with bronchitic symptoms and dyspnoea, but an association with heart disease was not demonstrated.