Increasingly, the nutritional impact of complex intervention packages is being sought. Typically, these complex interventions include intentional changes in the economic, social, and political environment as well as in medical and nutritional factors. This paper is concerned with the critical issue of the choice of indicators necessary to evaluate these complex interventions with multiple treatments. The authors first argue that large-scale, complex interventions require knowledge about indicator variables that is likely to emerge only from careful evaluations of simple interventions. They then discuss a set of optimal properties for field indicators of nutrition and health. These properties have implications for the design and evaluation of interventions. In light of these implications, the authors review the few evaluations of nutrition interventions reported in the literature in order to assess the knowledge now available for structuring complex, large-scale interventions. Finally, they explore problems of design and measurement peculiar to interventions with multiple treatment.