In order to define the amount of physical activity appropriate in primary prevention of coronary heart disease (CHD), we have compared the effects during 5 years of physical activity in four groups of middle-aged men with different but stable approximate metabolic costs of leisure time sports activities (AMCSA): sedentary (n = 40; 0 kcal per week), low activity (n = 31; 1-999 kcal per week), moderate activity (n = 56; 1,000-1,999 kcal per week), and high activity (n = 71; > or = 2,000 kcal per week). Time related increase of body mass and BMI was more pronounced in lower activity groups. Changes in HDL cholesterol were more favourable in the high activity group as compared to sedentary and low activity groups. The increase of diastolic blood pressure (DBP) in the sedentary group was statistically significantly different from the decrease of DBP observed in both moderate and high activity groups. We conclude that favourable long-term stabilization of most coronary risk factors is achievable with physical activity energy expenditure above 1,000 kcal per week. Physical activity-related energy expenditure > or = 2,000 kcal per week is associated with some additional benefits, especially with a favourable modification of HDL cholesterol level.