Description: On the basis of sociocultural theories of learning, expertise is not defined as a stable description of cognitive skills or formal ranks, but rather as a social role and nomination through the expert's constituency (Agnew, 1997). Accordingly, expertise is bound to social integration into networks of practice, in which they shape a professional identity (Lave & Wenger, 1998). Thereby the context of experts is rather a stable entity than a dynamic set of interrelations within and between activity systems of cultural derived rules and procedures (Engeström, 2004). This proposal aims to analyse the social embeddedness of experts within specific professional networks of practice, namely a scientific, coaching and medical network of experts. Within these networks, the interpersonal relationships as well as intrapersonal attributes are analysed in order to build up the interdependency and dynamics between expertise and context (Feltovich et al., 1997). Methodology: The proposal introduces a very young methodology of educational research, namely structural network analysis. The first step of analysis is a social network analysis, which focuses on relational data and investigates the context or the activity system of experts. A second study takes out the most central persons of the network and evaluates the egocentric networks of experts and their intentionality and agency in establishing and utilising their social resources (Palonen, Rehrl, Lehtinen & Gruber, in prep.). At least statistical studies about attributes of experts are conducted and combined with the quality of networks. Conclusions: Expected outcomes: Domainspecific results about the network quality of experts and its influence on expertise development. Social networks analysis is useful to represent the sociocultural context of domains and give background information about specific activity systems. The correlation with egocentric relations and the degree of expertise provide a deeper understanding of the context influence on professional learning.