The network success hypothesis predicts a positive relation between the networking activities of entrepreneurs and their economic success. Surprisingly, at first glance, empirical research on entrepreneurial networks and clusters in more than 30 reviewed papers offers quiet varying answers to the question if and why networks or clusters influence the development of start-ups positively. The dissimilar (or even contrary) empirical results might can be explained by differences in the theoretical models and the methods of measurement. An evolutionary model of the interplay between single entrepreneurs and their networks will be developed. Then, following and testing HAYEK’S (1971, 1979, 1988) institutional theory of group selection, it is argued that entrepreneurial networks which are build on a shared cognitive basis will create efficient rules of conduct with less effort, resulting in a capable institutional framework and thus being comparatively fitter systems.