Hospital studies were used to identify those characteristics of angina pectoris, cardiac infarction and intermittent claudication which most effectively distinguish these conditions from other causes of chest or leg pain. These are used to formulate precise definitions for epidemiological use and to form the basis of a standardized questionnaire.Agreement on the use of such a questionnaire would permit international comparisons of the prevalence of these conditions, as defined. This would not hinder the collection of additional information, as required in particular studies.As compared with physicians' diagnoses, the questionnaire had high specificity and reasonably good sensitivity. Interpretation of subjects' answers presents no serious difficulties. There is evidence that the diagnosis of angina pectoris presents special problems in populations with a high prevalence of chronic bronchitis.