The relevance of research evaluation in the knowledge society goes far beyond the measure of performances of researchers and their organisations. It matters a key factor of economy and society, seen by many scholars as a global public good: knowledge. As a consequence, research evaluation has relevant economical and social effects. However, the two most diffused techniques, i.e. peer review and bibliometrics, raise risks of perverse effects, particularly for sociological research, due to its peculiarities on which, moreover, still lacks a vast agreement. Some suggestions are put forward: a deeper insight into both these techniques, a strengthen of the publicity in both procedures and judgements, a careful monitoring of both publication and evaluation outputs, and also an appropriate attention to the third mission of universities and scientific institutions. All in all, research evaluation needs presently a vigorous impulse of research on its own methodological assumptions and actual applications. It comes out, therefore, the demand of a strong impulse to sociology of science that, in Italy, is far less developed then abroad. In particular, it seems opportune to encourage a sociology of sociology in order to regain the hoped authoritativeness of our discipline. In the coming knowledge-society all that becomes a pressing question.