In this case, we present a definition of survey data with multi-actor design as surveys in which several individuals who are related to each other by precisely defined social ties are involved. All individuals together with the well-defined social relationships and social roles define/constitute the multi-actor scheme. We explain also how the multi-actor design and the sample design are related to each other. The study ‘Divorce in Flanders’ is given as an example of survey data with multi-actor design. We discuss two options to structure the data: a flat-file database versus relational database management system. Since research questions with varying required level of analysis are typical for survey data with multi-actor design, we compare the possibilities to retrieve information from both types of databases. At last, we argue about the representativeness of survey data and how to apply general knowledge about adjusted weighting procedures in the context of multi-actor data.