During the past few decades, Bangladeshi children under age 10 have experienced significant improvements in nutrition, and sex differences in child nutrition have declined significantly regardless of family structure, a major change from previous observations in Matlab. However, our attempts to understand child nutritions in develop countries are hindered by problems with the measures used to evaluate health. The anthropometric proxies commonly used to judge nutrition (MBI, weight-for-age and height-for-age) often fail to capture the true health status of children in undernourished populations. Further the standard of comparison based on U.S. children misclassifies a large number of children in Banglades as malnourished, especially in the adolescent years. We explore nutrition in Matlab, Bangladesh, using measures of acute and chronic morbidity to assess whether and how anthropometric indicators of nutrition accurately reflect the health of children in this population.