This article opens up the blackbox through which evidence is selected and assessed in the making of guidelines and recommendations in global governance, through an exploration of “methods regimes.” Methods regimes are a special kind of sociomaterial arrangement, which govern the production and validation of knowledge, by establishing a clear hierarchy between alternative forms of research designs. When such regimes become inscribed in processes of global governance, they shape and control what knowledge is deemed valid and thus relevant for policy. We shed light that through a mode of operation that relies on a discourse of procedurality, a dispersed but powerful network of epistemic operators, and a dense web of infrastructures, methods regimes constitute and police the making of “policy-relevant knowledge” in global governance. Through an examination of the case of “GRADE” (Grading of Recommendations, Assessment, Development, and Evaluation), a standardized system that evaluates and grades the quality of evidence in global health, we show that its dominance has worked to the effect of empowering a new cast of methodologists, seen as more objective and portable across domains, sidelining certain forms of evidence that do not conform with its own methodological criteria of scientificity, and “clinicalizing” research in medicine and beyond.