This paper presents a recent review of the policy networks literature, both empirical and theoretical, and describes an ongoing project to map existing policy network data. Much of the literature on policy networks sees them variously as a mode of governance, and a metaphor for the reality of governance, or configurations of individuals/organisational engaged in a policy sector (Rhodes 1992b). Dowding asserted that the field of policy networks is divided into those who believe that the network gives rise to a policy through convergence (ties forming through mutual beliefs and values within an epistemic community – the advocacy coalition framework (Haas 1992, Sabatier- & Jenkins-Smith) or those who perceive policy networks as a way of linking state-level structures with interpersonal ‘micro’ interactions; i.e. a description of how entitites are linked through the enacting of power and bargaining (Harsanyi, Rhodes & Marsh). Therefore, to be analytically useful, the concept of a policy network must be able to explain in network terms (structure, position, characteristics) policy outcomes or processes. These explanations must work in a range of domains and contexts. Understanding policy network formation, change and process can inform us not only about how policy processes work, but also how ideas and evidence are found and passed between actors. The theory of policy networks should be able to explain the common features of policy networks (as opposed to networks of other kinds), which this project is aiming to identify. Using an interdisciplinary, multi-theoretical and methodological approach, our project will add to the conceptual and empirical literature on policy networks. We end with some proposals for the development of the field, and invitations to participate in the ongoing project.