Social activities are responsible for an important share of trips conducted by individuals. This paper contributes to a rapidly increasing stream of transportation research into individuals' choice of social relations and trips made to maintain their social relations. A method that was used to collect data on ego-centered social networks in a large-scale survey conducted in the Netherlands in the late 1980s is described, and a modelbased framework for analysis of these data is proposed. The framework consisted of a set of linked regression and choice models for analysis and prediction in a coherent fashion of the travel demands resulting from a chain of decisions, including the size and composition of the social network and the travel time and contact frequency related to each member of the network. The results indicate that the sociodemographic attributes of an individual have only a modest influence on network size and a stronger impact on the choice of relationship type, travel time, and contact frequency. Furthermore, significant relationships exist among these social network variables. It was concluded that the framework provides a straightforward and useful way to analyze social network data and identify fruitful directions for future research.