Work on knowledge infrastructures of the past decade has shown that infrastructures are central to many crucial types of knowledge about climate change, human migration or biodiversity. They are also important in supporting diversity and pluralism in knowledge production. But how can knowledge infrastructures best be evaluated?