Since quite some time, at least in the United Kingdom, a campaign against complementary and alternative medicine (CAM) treatments and training courses has been happening. Given some prominence in the media, this campaign seems to aim at having CAM courses scrutinized, CAM treatments taken off reimbursement schemes, and brandishing doctors and academics dealing with CAM as quacks. The buzzword used is "evidence-based," with the presupposition that, while conventional medicine is "evidence-based," CAM is not. It is worthwhile to examine the notion of "evidence" used in this discussion. I argue that it is a rather unwholesome type of evidence that is referred to here, namely only the difference between an active treatment and a sham treatment. For several reasons this is too restricted and not sound, as it creates several paradoxes, for instance the efficacy paradox. This consists in what is now empirically documented, whereas a CAM placebo can be actually more effective than a conventional, supposedly evidence-based, treatment. This shows that the discussion is really about something completely different: CAM has become a real player in the health services field, threatening revenues of the pharmaceutical industry. Taken together with a series of events this suggests that this campaign is carefully staged. It requires that we take the question of what "evidence" means to the heart of the methodological discussion, also for the sake of a broader, more wholesome notion of what counts as evidence. It also requires that we keep producing good data with a broad enough outlook comparing CAM treatments with conventional ones, and not just with placebo.