To determine whether body mass index (BMI, kg/m²) or percentage body fat (%BF) by bioelectrical impedance analysis (BIA) better reflects the cardiovascular risk profile, we examined the associations among BMI, %BF by BIA, and cardiovascular risk factors (systolic blood pressure (SBP), diastolic blood pressure (SBP), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), LDL-C/HDL-C ratio, and triglycerides (TG)) in 1, 217 Japanese male office workers aged 25 to 59 years. From stepwise regression analyses of cardiovascular risk factors on age, BMI, alcohol intake, and cigarette smoking, significant correlates were, in the order of relative importance: age, BMI, and alcohol intake for SBP and DBP (the cumulative percentage of variation; 14.9% and 21.3%, respectively); age, BMI, and alcohol intake (negative) for LDL-C (11.0%); BMI (negative), alcohol, and cigarette smoking (negative) for HDL-C (19.9%); BMI, alcohol intake (negative), age, and cigarette smoking for LDL-C/HDL-C ratio (23.1%); and BMI, age, cigarette smoking, and alcohol intake for Log TG (21.7%). From stepwise regression analyses using %BF by BIA as an independent factor, %BF by BIA was also significantly associated with each cardiovascular risk factor, but the decrease in explained variance for each cardiovascular risk factor was 0.2-4.5%, compared with the model using BMI as an independent factor. These results suggest that BMI may better reflect blood pressure or serum lipid profile than %BF by BIA.