The purpose of this paper is to develop the micro foundations of a theory of innovation systems by exploring patterns of interaction between innovating firms and a set of actors in a Dutch region. First, two largely detached literatures on the interactionist perspective on innovation are described. There is the theoretical literature of Lundvall, and the empirical literature of Von Hippel and Pavitt. Lundvall's theory stresses the importance of firms and structured interaction among firms for the innovation process. While the empirical literature stressed the importance of the users in the innovation process, as well as internal departments like R&D, sales and production, and finally basic science. Connecting both literatures produced an actor set involved in the innovation process which is much broader than Lundvall's, and Lundvall provides the hypotheses about the antecedents of interaction among the actors involved in the innovation process. Basically Lundvall argues that patterns of interaction are affected by the the nature of the innovation process. By means of separate indicators for the nature of the innovation process, and separate variables measuring patterns of interaction we reworked Lundvall's hypotheses. This allows us to explore the regional system of innovation of North-Brabant from within, answering the following research questions: To what extent do suppliers, customers, the knowledge infrastructure and intermediate organisations differ in their contribution to the innovation process of innovative firms? To what extent are patterns of interaction affected by technological dynamics? Our findings revealed that patterns of interaction are indeed contingent on the nature of the innovation process. The type of actors involved in innovation processes were found especially within the value chain customers and suppliers - and besides that inside the firm mainly production, sales and R&D. Also we found that the intermediairy organisations like trade organisations, chambers of commerce and higher ...