Study objective - The aim was to investigate the associations between physical fitness, leisure physical activity, and coronary risk factors. Design - This was a cross sectional study of a random sample of men and women, following a population survey. Setting - The municipality of Tromso, Norway in 1986-1987. Participants - All men born 1925-1966 and all women born 1930-1966 were invited to the survey; 21 826 subjects attended (81% of the eligible population): of these, 297 men and 312 women, randomly selected, attended the present study (attendance rates 94% in men and 89% in women). Measurements and main results - Fitness was tested by bicycle ergometry. Physical activity was reported on a questionnaire. Multiple regression analysis was performed with fitness and leisure activity as dependent variables, and coronary risk factors as independent variables. Fitness and leisure activity were positively related (p < 0.05). Prominent findings for fitness were negative associations with age and smoking (p < 0.05), and positive associations with body mass index in both sexes (p < 0.01). HDL cholesterol and systolic blood pressure were significant predictors of fitness in men (p < 0.01). Smoking emerged as a strong negative predictor for leisure activity in women (p < 0.01), and a negative relation between leisure activity and total cholesterol was found in men (p < 0.01). Conclusions - The study indicates that coronary risk factors are more closely linked to physical fitness than to leisure physical activity.