Quality and impact are increasingly the two pillars of research evaluation. These categories are often taken for granted as attributes of research that can be measured and judged in a rational assessment process. But how do these evaluation processes actually work? The chapter argues, on the basis of a summary of key results from the scholarly literature, that quality and impact are actually the result of, rather than the input into, the evaluation activities. Both can be interpreted as boundary objects at the interface between the world of research and the world of policy.