The development of obesity in childhood is considered a major determinant of cardiovascular risk. Currently the body mass index (BMI = weight/height(2)) is widely used as a measure of obesity. However, since BMI is associated with height during childhood, a weight for height index (weight/height(p)) that is independent of height is thought to be more appropriate. Therefore, to compare the utility of such weight/height(p) index with BMI in assessing adiposity and its relation to cardiovascular risk variable data from the Bogalusa Heart Study participants aged 6 months to 21 years were examined. A total of 31,796 observations on 12,827 subjects was used in the data analysis. Study variables include height, weight, subscapular and triceps skinfolds, blood pressure, serum lipids and lipoproteins, and plasma glucose and insulin. The optimal exponential for the weight/height(p) index started from 2.42 in the 6 month olds, decreased to 1.86 in 2 to 3 year olds, increased to 3.29 among 10 to 11 year olds, and then decreased to 2.15 in the 20 to 21 year olds. The BMI showed slightly higher correlations than weight/height(p) index with subscapular skinfold in children. Both in children and young adults BMI also showed a slightly higher correlation with other cardiovascular risk factor variables regardless of age-race-sex groups. These results indicate that weight/height(p) index is not superior to BMI as an indicator of adiposity and related cardiovascular risk factors during childhood.