In this article, the authors discuss conceptual and pragmatic considerations for conducting research of families using large secondary data sets. Conceptual considerations include establishing consistency among the theory, variables, and available data, and determining reliability and validity of the data in the context of the theory. Pragmatic considerations include the use of resources such as management of the data among several authors, criteria and methods for selection of a subsample, and, recoding of the data to examine dyadic difference scores. The Family Special Interest Group of the Eastern Nursing Research Society initiated this research as part of a project to analyze families using large national data sets. The purpose of the secondary analysis was to identify family beliefs about healt- promoting behaviors. Combining parent and teen data to create relational level data resulted in new information that had not been identified in the original survey.