OBJECTIVE:To investigate whether the favourable cardiovascular disease (CVD) risk factor profile of habitual exercisers is attributable to exercise or leanness. DESIGN: Cross-sectional study of 113 nonsmoking men aged 30-45 y. CVD risk factors were compared in exercisers (n ¼ 39) and sedentary men (n ¼ 74), and in subgroups of lean exercisers (n ¼ 37), lean sedentary men (n ¼ 46) and obese sedentary men (n ¼ 28). Waist girth was used to identify lean (o100 cm) and abdominally obese (Z100 cm) subgroups. MEASUREMENTS: Blood pressure, physical activity (7-day recall), physical fitness (maximum oxygen consumption) and fasted lipoproteins, apolipoprotein (apo) B, triglycerides, glucose and fibrinogen. RESULTS: Exercisers were fitter and leaner than sedentary men and had a better CVD risk factor profile. Total cholesterol, LDLcholesterol and apo B concentrations were lower in lean exercisers than in lean sedentary men, suggesting that exercise influences these risk factors. Indeed, time spent in vigorous activity was the only significant predictor of total cholesterol and LDL-cholesterol in multiple linear regression models. Exercise status had little influence on triglycerides and HDL-cholesterol (HDL-C), and unfavourable levels were only evident among obese sedentary men. Waist girth was the sole predictor of triglycerides and HDL-C, explaining 44 and 31% of the variance, respectively. CONCLUSIONS: These findings suggest that the CVD risk factor profile of habitual exercisers is attributable to leanness and exercise. Leanness is associated with favourable levels of HDL-C and triglycerides, while exercise is associated with lower levels of total cholesterol, LDL-cholesterol and apo B.