Nutritional indicators are used to screen, diagnose, and evaluate interventions in individuals. They are also used in populations to ascertain and place under surveillance the magnitude of nutritional problems, their location and causes, and to evaluate the impact of programs and policies. Nutritional indicators are also used for research to make inferences about biological and social mechanisms affecting or being affected by nutrition. All these activities include measurements of nutritional indicators, but the choice of indicators, their measurements, analyses, and the need for other data can be very different for inferences from research, for patient management, for making public policy, or for planning or evaluating programs. There is no best indicator, best measure of an indicator, or best analysis of an indicator in a generic sense. The definition of "best" depends ultimately on what is most appropriate for the decision that must be made. This paper gives examples.