Leon, A. S. (School of Public Health, U. of Minnesota, Minneapolis, MN 55455), D. R. Jacobs, Jr., G. DeBacker and H. L. Taylor. Relationship of physical characteristics and life habits to treadmill exercise capacity. Am J Epidemiol 1981; 113: 653-60. Apparently healthy middle-aged men (n = 175) were recruited from a population sample and completed questionnaires about habitual physical activity, smoking, beverage consumption and sleep habits. Body mass Index (BMI), heart rate and blood pressure were measured at rest and during submaximal exercise; frequency of ventricular premature beats (VPB) on an ECG rhythm strip; hand grip strength; and serum cholesterol. These characteristics were correlated with duration of treadmill exercise by the Bruce protocol. Univariate analysis Indicated that treadmill performance was significantly and positively correlated with leisure-time physical activity and personal reports of sweating and/or dyspnea occurring regularly during such physical activity. Performance was negatively correlated with age, BMI, resting heart rate, cigarette smoking, and consumption of caffeine-containing beverages, but was Insignificantly related to job physical activity, hand grip strength, alcohol consumption, sleep habits, blood pressure, cigar smoking, serum cholesterol, and the frequency of VPB. A 0.75 multiple correlation coefficient was found between treadmill performance and 11 of the above variables and the r is increased to 0.81 by adding heart rate during submaximal exercise. It is concluded that substantial prediction of work capacity and physical fitness of population is achieved by questionnaires and easily obtained, noninvaslve physical measures.