This paper examines the context of evidence-informed practice (EIP) by inquiring into how educational practice is defined and organised, and how predominant understandings of educational practice are concomitant with preferences for particular forms of evidence. This leads to discussion of how certain educational research traditions speak (or are unable to speak) to these evidence requirements, and how this shapes the nature of EIP. While the rise of EIP can be understood as part of the increasing attention paid by governments to systemic 'improvement' in education systems, it can be argued that the lack of a coherent body of educational knowledge in many national traditions enables governments to exercise control not only of definitions of 'what works' in education but also over conceptualisations of educational practice. For some policy makers and practitioners, the much-remarked dislocation between 'evidence' and teaching practice in many national contexts can only be solved by a narrowing of what counts as knowledge alongside a more prescriptive control over what counts as acceptable educational judgement. However, such an alignment serves to exclude wider educational purposes and arguably instrumentalises pedagogical relations. Meanwhile, some continental European countries maintain traditions that may serve to mitigate such developments, although these traditions are not without challenge.