This study was motivated by a desire for a deeper understanding of why Filipino infants are undernourished so early in their lives, and what might be done to prevent this. Data from an urban poor Philippine community was used to conduct an analysis of complementary feeding (CF) practices and behaviors from two different perspectives: the biological and the sociocultural. The underlying structure was provided by a global situation where scientifically-based guidelines for CF had been framed, and a set of indicators, to track progress in adherence to these guidelines and achievement of policy and program goals based on them, were in the process of validation. The over-all objective was to determine how the main CF indicators - Diet Diversity (DD) and Meal Frequency (MF) - would perform in assessing the status of infants 6 - 11 months old , and how this kind of assessment might ultimately be of use. Ethnographic techniques were used to investigate the conceptual agreement between nutritional/public health professionals and mothers of the study infants. Definitional issues about foods and liquids, mothers' perceptions of breastmilk as a unique part of the infant's diet, and local concepts about meals and snacks and breastmilk were identified that have the potential to generate inaccurate communication as well as introduce a problem for DD and MF construction. This study demonstrates why local adaptation is essential. DD is related to growth, but not the individual food groups; adding MF weakens this predictive association. The rationale for DD's use as an indicator of diet quality is its relationship with mean micronutrient adequacy(MMDA). MMDA was found inadequate even at the highest DD score. Intake of fortified products led to an attentuation of the relationship between MMDA and DD. The contributions of each food group to MMDA with each unit increase in DD is the same for breastfed (BF) and nonbreastfed (NBF) infants if breastmilk is counted as a food group. Not accounting for breastmilk's contributions to DD, as is the current practice, is conceptually counter-intuitive and potentially confusing requiring separate cut-offs for the BF and NBF.