Skinfold anthropometry is performed world-wide with little appreciatio
n of its limitations. This paper shows that it is important to appreci
ate the magnitude of the biological component of the residual error in
any regression of body density against sums of skinfolds; for example
, for young adult males a biological error of around 21% (s.d.) is obs
erved. Such 'errors' simply express nature's refusal to conform to imp
osed regression relationships, reflecting the inherent variability in
the ratio of Subcutaneous to total body fat. (C) 1998 Elsevier Science
Ltd. All rights reserved.