The aim of the present study was to determine the relationship of BMI with other cardiovascular risk factors, leisure-time physical activity and diet. Participants were recruited in a cross-sectional population-based survey in a southern-Europe Mediterranean population (Spain); cardiovascular risk factors were measured, and leisure-time physical activity and diet intake were evaluated. Linear regression analysis adjusted for several confounders showed a significant, direct association of BMI and total cholesterol (P<0.005) and LDL-cholesterol (P<0.006), in men. HDL-cholesterol was inversely related to BMI in both sexes (P<0.0001). Higher BMI was more frequent in less-active men (P<0.04) but not in women. BMI increased significantly (P<0.0001) by 1.92 kg/m(2) with each 4.18 MJ consumed in men but not in women. Dietary intakes of carbohydrate (P<0.03), total fat (P<0.03) and saturated fatty acids (P<0.02) were directly associated with BMI in men but not in women, in whom protein intake was correlated (P<0.001) with BMI. Linear regression models including dietary components explained up to 10.6 and 21.1 % of BMI variability in men and women, respectively. Sex differences in the association of BMI with total cholesterol, and LDL-cholesterol, may account for the lower risk for CHD in women compared with men of similar BMI reported in the literature for the southern-Europe Mediterranean region. An increases of BMI may be more deleterious in populations in which it is accompanied by other risk factors such as a higher intake of total fat and, particularly, of saturated fatty acids, or lower leisure-time physical activity.