Rationale: The new area of health system research requires a revision of the taxonomy of scientific knowledge that may facilitate a better understanding and representation of complex health phenomena in research discovery, corroboration and implementation. Method: A position paper by an expert group following and iterative approach. Results: ‘Scientific evidence’ should be differentiated from ‘elicited knowledge’ of experts and users, and this latter typology should be described beyond the traditional qualitative framework. Within this context ‘framing of scientific knowledge’ is defined (...) as a group of studies of prior expert knowledge specifically aimed at generating formal scientific frames. To be distinguished from other unstructured frames, FSK must be explicit, standardized, based on the available evidence, agreed by a group of experts and subdued to the principles of commensurability, transparency for corroboration and transferability that characterize scientific research. A preliminary typology of scientific framing studies is presented. This typology includes, among others, health declarations, position papers, expert-based clinical guides, conceptual maps, classifications, expert-driven health atlases and expert-driven studies of costs and burden of illness. Conclusions: This grouping of expert-based studies constitutes a different kind of scientific knowledge and should be clearly differentiated from ‘evidence’ gathered from experimental and observational studies in health system research. (shrink)