The process of consultation has become integral to the development, implementation and evaluation of a raft of UK health and social policies. However, the current bewildering patchwork of area-based initiatives means that, in many localities, it is impossible to evaluate the outcomes of particular targeted initiatives, let alone make sense of local planning consultations, Best Value reviews and (multi-agency) service reviews which run concurrently. The cumulative effects of this consultation "overload" threaten to swamp both local authorities and their service users. Consultation is itself a crucial yet deeply problematic process. There is an official view which holds that an "old" model of consultation - often tokenistic and unrepresentative - is being replaced with a " new" one. This paper examines and challenges that view in relation to the key policy areas of housing, social services and policing. it also pays particualr attention to, and problematizes, the notion of "hard-to-reach groups", which is so dominant in the discourse of consultation. The paper argues that developing approapriate tools and recognizaing that consultation is a process - not an event - are essential starting points in addressing these problems. The next step is to reconscile the principles of both evidence-based policy and user-led services into a strategic (and "joined-up") framework. But, when all this is accomplished, we still need to question the political and fiscal contexts in which policy-making takes place and within which the process of consultation is itself bounded