The study of clusters is an important area of academic research. Clusters are also an important tool for public sector economic development. Historically, many organically developed clusters were based around a community of like-minded people. This may have led to a commonly held assumption in the literature that all clusters are communities. We suggest that not every cluster is also a successful community, based on empirical findings for four biotechnology clusters. Three issues of interests to practitioners are causes for clusters not being communities, the link between cluster success and development of ocommunityo and policy failure at creating community within clusters. The discussion also aims to extend theory, first, by dealing with the rarely discussed question of the lack of collaboration oacrosso communities, and secondly, by attempting to help develop a theory of community formation in clusters.