Despite increased interest within Translation Studies to provide orientation for translation quality assessment (TQA), academic efforts in this area are still largely ignored, if not explicitly rejected by the profession. The purpose of this paper is to investigate why scientific models for evaluating translations are difficult to apply and to outline a number of ways in which the gap between theoretical approaches and practical needs may be negotiated. Following a critical analysis of some TQA models, the paper suggests that a reductionist view of translations as products and a neglect of the conditions under which translations are produced ultimately result in evaluation criteria which cannot account for the individuality of target texts.