This study has been conducted to find out whether nutrition indi­cators can assist in identifying and measuring poverty, and particularly, in finding out whether the most commonly used nutrition indicator, based on anthropometric measurements in children, is a useful proxy index of differences in socio-economic conditions in the poor. Data on the growth of Japanese children collected over the last 25 years has been used to establish changes in nutritional status and then these patterns have been compared with other development indi­cators. This time sequence of change has been compared with cross- sectional data from low income groups in four South-East Asian countries (Indonesia, Thailand, Malaysia and Singapore) in 1975, and they covered a wide range of per capita GNP. A set of nutrition indicators (prevalence of anaemia, weight-for-age deficits, and 2-5-year mortality rate) have been compared with a welfare and an economic indicator. The data for nutrition indicators were from surveys which used the household as a sampling unit. The validity of an anthropometric measurement (weight-for-age) has been tested in terms of sensitivity and specificity as a screening test of the nutritional status of children. The change in the nutritional status of children in Japan is greater during the early stage studied than in the later stages. Other development indicators also show continuous improvement over time but the rate of change differs from that shown by the nutrition indicator at equivalent stages. The nutrition indicator demonstrates that urban-rural disparity disappeared at an early post-war period. The patterns of a nutrition indicator based on weight measurements and of other nutrition indicators from cross-country data are similar to those of the welfare indicator. However, the study shows the potential value of a nutrition indicator as a reflection of the living standards of the poor from both the cross-country and the longitudinal observation.