It is necessary that it has circulation for the interior of the services and not only the use of theories or ' ' technique-pura' ' (Amarante and Carvalho, 1999), thus making possible the interaction between the researcher and the institution. When we think about the possible pointers to be enclosed in this process of evaluation of a service of mental health and the occured attempts already in some services of three states of Brazil (Minas Gerais, Rio De Janeiro and So Paulo), we perceive that all point with respect to the difficulty of if mensurar changeable, a time that these bring explicit and implicit factors as motivation of the diligent professionals, expectations and satisfactions of the team with the service, number of professionals contracted for that service, horria load, among others. We consider despite to adopt pragmatic avaliativos criteria for this process, it is not enough to enclose the complexity of the experience of a specific public as the carrying users of suffering mental. Finally, let us consider the importance to exist a definition of a theoretical model as previous stage the construction and public implantation of programs and services, what it finishes not happening as part of projects in our country: (…) an evaluation process can also include the construction of a theoretical model of the reality to be studied, from existing the implicit and explicit objectives, in legitimated way that the end item is argued and for the involved social actors. It can be affirmed that pra to carry through any type of evaluation is necessary that they are available pointers capable to quantify and/or to characterize of the form the next possible to the Real facts, processes and complex situations that cannot be apprehended without mediaes. Of this form, ' ' (…) the pointer is, simply, a form of approaching in them to the knowledge of that we cannot catch diretamente' ' or then we can understand it as ' ' (…) specific measures (explicit) and objective verifiable of the changes or resulted of one atividade' ' (AGUILAR; ANDER-EGG, 1994, pp.