In the widespread complaints about the prevalence of party influence in local government there is a recognizable stereotyping of local politicians and parties which is a caricature of the complex reality. The purposes of this paper are to question the validity of the stereotypes and to suggest the abandonment of simplistic categories in favour of categories which go some way at least towards recognizing the complex variables which shape the role-playing of elected members. What follows is based on a ten year period of participant-based research in an urban county borough council. Material from this specific study was supplemented by participant observation in a new county council and by observation-based research in two other boroughs. Participant research is a technique which creates a number of problems, the most formidable being that the researcher is a ‘partisan’ role-player in the ongoing local authority activities. In Halifax (the first-mentioned urban authority) this was not a great problem because the context was influenced by a three-party system in which the ongoing decisionmaking was tripartite and, indeed, occasionally multi-partite.