It has become common public practice to consider the nutritional statu
s of children under five years (as measured by anthropometry) as a bar
ometer for the population as a whole, even though no published literat
ure confirms the relationship between the nutritional status of differ
ent demographic segments of the population. This study investigates th
e relationship between maternal and child anthropometry using data fro
m a regional cross-sectional survey recently undertaken in the Republi
c of Guinea. The degree of correlation between maternal and child anth
ropometry is assessed and the performance of child anthropometric stat
us as a screening tool for maternal anthropometric status is evaluated
using various epidemiologic techniques. The major finding of this stu
dy is that although maternal and child anthropometric indices includin
g receiver operator characteristic (ROC) analysis are significantly co
rrelated, the magnitude of correlation is modest. More importantly, ma
ternal malnutrition exists in many households where children are free
of malnutrition and vice versa. The findings of this research suggest
that the common practice of limiting nutritional assessment to under-f
ives may result in important inaccuracies in evaluating community nutr
itional problems.