This article examines the use of the concept of "evidence" in the current movement known as "Evidence-Based Medicine" (EBM). The article includes some historical milestones of the concept, in which Aristotle, Andr & eacute; Lalande and Michel Foucault are the main references. Then, the process through which EBM was born is reviewed, going through Cochrane, Guyatt, Sackett and Woolf. It is explained that EBM understands evidence as facts, where controlled and randomized clinical studies, as well as systematic reviews, are the most appropriate means to obtain evidence. We propose that EBM may not be aware that the concept of "evidence" used is crypto-normative, which makes this concept contain a strong axiological load. The article does not intend to simply dismiss the concept of "evidence", but rather to review its univocality and open it to an analogous understanding that could improve clinical practices and the theoretical scope that serve as their foundation. A central part of the arguments used in this paper consists of reviewing the supposed epistemic neutrality of EBM, which is a movement in formation that requires various adjustments. Among them, an in-depth review of the concept that gives it its name and meaning, for the benefit of the generators and users of EBM (clinicians and patients).