This paper examines propositions of this model using data from a 1984 study of a Toronto policing project designed to encourage a flexible, planned approach to local policing. Telephone interviews with residents of 2 Toronto communities (458 and 451 residents, respectively) examined attitudes toward and experiences of crime, order, and policing problems. Survey results failed to provide Toronto police with a clear indication of an unambiguous set of community standards or concerns and failed to validate the policing model assumption of value consensus. The reform model is riddled with ideological rhetoric about the nature of the contemporary community that is at odds with findings. Further, the organizational imperative of conventional policing demands a hierarchical command structure that is necessarily in conflict with the horizontal nature of local communities. Police administrative districts bear little resemblance to natural communities. There was, however, a surprising similarity in police and resident rank ordering of perceived community problems, suggesting that the Toronto Police were already allocating resources on the basis of citizen demand and were aware of significant neighborhood crime problems. This suggests that conventional policing is to some extent citizen and community based, and that community-based policing is in fact a reformulation or modification of the traditional policing model. The propositions of community policing remain largely untestable because of unresolved definitional issues surrounding the concepts of community, public order, and community consensus. 4 tables, 2 notes, and 25 references. (Author abstract modified)