The association between visceral adipose tissue (VAT) with cardiovascular disease (CVD) has been clearly demonstrated. Although typical VAT area at 4th and 5th lumbar vertebrae (L4-L5) is used to approximate VAT volume, growing evidence has suggested that this measurement site may not be ideal. However, these findings for Asian people remain unclear. Thus, we searched for the better VAT measurement sites associated with CVD risk factors in obese, Japanese men. Eighty-two obese men were included in a cross-sectional study. Among these participants, 37 men completed the 12-week intervention (90 min and 3 d/week) were used for addressing longitudinal association between the VAT measurement sites and CVD risk factors. Consecutive MRI images (from 3 cm below L4-L5 to 20 cm above L4-L5) were used to explore the relationship between each VAT area and CVD risk factors (total cholesterol, HDL cholesterol, triglycerides, glucose, insulin and blood pressure). The images located only 5-9 cm above L4-L5 had significant correlations with HDL cholesterol and triglycerides, but L4-L5 site did not in the cross-sectional analysis. In response to exercise, the image located 5 cm above L4-L5 showed the highest correlations with changes in total cholesterol (r = 0.46) and glucose (r = 0.36). Also, the image located 6 cm above L4-L5 showed highest correlations with changes in triglycerides (r = 0.37) and insulin (r = 0.37). Thus, the range of VAT images located 5-6 cm above L4-L5 may be optimal for identifying CVD risk factors compared to a typical site of L4-L5.