Regional variability of skinfold thickness differs from individual to individual. This may be affected by differences in total body composition. If so, individual's fat distribution patterns may contribute to the prediction of body composition. On this premise, the regression relationships between body composition and fat distribution patterns which is represented by the pair-wise contrast of fatfolds, were analyzed by the quantification method of the first type. This method is a multivariate analysis method for categorical data. First, skinfold thickness was measured at 9 sites. From these 9 sites, independent pair-wise combinations of measures were established, for example subscapular>abdomen. Within each pair-wise contrast if A>B then the statement is true and a value of 1 was attached. If the statement is false a 0 was recorded. Thus the pair-wise relationships of skinfold measures were expressed as categorical data. These data subsequently served as predictor variables for multivariate regression analyses with the dependent variable being fat mass. The results suggested that regardless of gender, the series of pair-wise contrasts of fatfolds significantly correlated with fat mass. However, the reliability of the equations is not sufficient for the prediction of body composition.