Estimates on the prevalence of undernutrition at the level of regions in the world are widely divergent, whether assessed on the basis of food energy inadequacy or on the basis of anthropometry. In this study a comparison between the two approaches in the assessment of undernutrition has been made at the level of individual countries, utilizing until recently unavailable anthropometric data for adults. Results indicate that the two approaches yield strongly divergent and even contradictory results. In view of the complex methodology and tremendous data needs for estimating prevalence of food energy inadequacy and the relative straightforwardness of anthropometric assessment, it is concluded that anthropometric data are likely to be a more reliable source of information for estimating prevalence rates of undernutrition. Findings call for a re-assessment in current practices with respect to measuring and reporting undernutrition. © 2001 Elsevier Science Ltd. All rights reserved.