Pain is a refractory problem in healthcare. One proposed solution is to promulgate best practices derived from the highest quality evidence stipulating how to assess pain and which treatment interventions are the most efficacious. Such a solution seamlessly integrates into evidence-based medicine, the dominant ‘revolutionary paradigm’ in healthcare. This paper argues that such a ‘revolutionary’ approach is theoretically untenable and ethically abhorrent. Section 1 adumbrates the evidence-based medicine framework, how best practices are often formulated under its auspices, and which methods it might use to address pain. Section 2 points out that this approach faces profound theoretical issues when applied to pain, even by its own lights. Section 3 argues that attempts to use such best practices in the clinic generate grave ethical problems. Finally, Sect. 4 briefly sketches a different form of medicine, one that casts evidence as a helpful handmaiden of medicine, instead of a domineering mistress.