During 1968 to 1970, approximately 17 000 middle-aged male executive grade civil service officers, all of them engaged in sedentary or very light work, recorded on a Monday morning their leisure time activities over the previous Friday and Saturday. In 1971 a sample of 509 of these men completed further questionnaires for medical, social, and smoking history; these men had a resting electrocardiogram, and height, weight, skinfold thickness, blood pressure, and plasma total cholesterol were measured. Vigorous exercise in leisure time had previously been reported by 125 (25%) of the men, and these as a group had significantly fewer electrocardiographic abnormalities (changes compatible with myocardial ischaemia, ectopic beats, and sinus tachycardia) than the men not reporting vigorous exercise (P less than 0-02). This difference remained when all men with any history suggestive of cardiovascular disease were excluded from the analysis. Blood pressure, plasma total cholesterol, and smoking habits were examined with respect both to vigorous exercise and to the electrocardiogram, but the only relation found was that electrocardiographic abnormality increased with increasing blood pressure. Even among men with higher pressures, however, those reporting vigorous exercise had fewer electrographic abnormalities than the others. The results provide further support for the association of habitual physical activity with coronary health.