Five evaluations of the same community regeneration project in Newcastle upon Tyne are compared. It is argued that the separate evaluations are weak in method, atheoretical, ahistorical and lacking in a sense of social structure. To progress, community evaluations must typologise communities rather than treat community as a nebulous quality of relationships. A typology is offered. Further, it is suggested the community regeneration in this case was an extension of urban governance which artificially constructed what is called an ‘inverse’ community.