Skinfold thicknesses represent a comparatively simple and reasonably a
ccurate assessment of body fatness which is an important part of the e
stimation of nutritional state. However, much emphasis is placed on th
e necessity to be very precise in the exact position of the skinfold b
eing measured and there is frequently concern, also, about the variabi
lity of the measurement by different observers using different types o
f caliper. Fifty-three women and forty-five men had four skinfolds (bi
ceps, triceps, supra-iliac and subscapular) measured first at the stan
dard sites (Tanner, 1953; Edwards et al, 1955), and then at deliberate
ly chosen sites about 20 mm distant from the 'correct' ones. The effec
t on the estimation of body fatness by this manoeuvre resulted in a di
fference of usually less then 1 % and, at a maximum, 3 %. In the light
of the uncertainty of the basic assumptions which are made in extrapo
lating from skinfolds (or indeed densitometry, total body water, total
body K, and others) to body fatness, these technical errors assume co
mparatively little importance.