This paper addresses the question "How does practice generate valid research outcomes?". The question is of general significance to all vocation-oriented disciplines (including Design) as they move to integrate more fully within the established academy. It is of significance, not least, because the established academy is already well versed in recognising and rewarding research undertaken within the scientific and (to an extent) the humanities traditions, but largely fails to acknowledge the kinds of activities vocational academics typically view as being central to their own scholarship. There is a mismatch between what the established academy views as appropriate research outcomes and the kinds of activities viewed as relevant to practice-based disciplines. Attempts to rectify this mismatch (in the Australia context at least) have focussed primarily on equating the research outcomes generated by practice-based disciplines with traditional research outcomes. The results so far have been unsatisfactory on all counts. It transpires that the kind of problems and problem solving typical of a professional practice setting are very different to the problems for which science has proven to be such a powerful epistemological base: the practice situation is complex, uncertain, changing, unique, and often ranges across ostensibly incommensurate value structures. Rigorous science requires such problems to be restructured and preprocessed, and it is in that prior restructuring that the relevance of the research outcomes tend to be abstracted (and distanced) from the immediate concerns of practice. This paper argues an alternative approach. The alternative approach is to develop an epistemology of practice itself: an approach which recognises practice as a valid form of knowledge creation within itself. It is apparent that design, as a generic process, is itself particularly well suited to just the problem context faced by practice in general. This paper will argue that design itself provides the basis for a new form of epistemology particularly relevant to practice. We distinguish the new epistemological base from practice per se ("mere" practice) through the notion of "scholarly design." Scholarly design challenges the practice-based researcher to determine a legitimate scope for each study, test the claims being made, present the knowledge it generates in an appropriate form, have utility to practice, and guard against relativism. These questions will be further addressed in the paper.