The past twenty years have seen a remarkable paradox in popular attitudes to medicine. While regulations for licensing drugs and appraising doctors have become increasingly stringent, more and more patients have turned to untested therapies supplied by unregulated practitioners. An increased demand for evidence-based medicine has been paralleled by a steady rise in therapies based on tradition and opinion. This divergence cannot be explained by two populations of patients: the evidence is that the same patients use both approaches. Patients considering any new treatment want to know if it works and if it is safe, but seem to have two different sets of rules. The divide is both linguistic and philosophical: terms such as ‘natural’, ‘traditional’ and ‘holistic’ are difficult to compare with confidence intervals and P values. So what happens if we judge both types of medicine by the same standards? Does complementary and alternative medicine (CAM) have a valid place alongside orthodox medicine? The Desktop Guide, originating edited from the University of Exeter Department of Complementary Medicine, takes a serious look at the evidence for efficacy and safety of CAM. The opening section on investigative methods sets the tone of clarity, transparency and depth that characterize the book. Section 2, which summarizes diagnostic methods from bioresonance to the Vega-test, is informative and helpful. For each technique, published studies are systematically reviewed with verdicts ranging from ‘the method is not valid’ (iridology), through ‘its value as a diagnostic tool seems limited’ (Kirlian photography), to ‘some but not all diagnostic methods are valid’ (chiropractic). Section 3 takes us through therapies from acupuncture to yoga, and section 4 herbal and non-herbal medicine from aloe vera to yohimbe. Information including constituents, rationale, pharmacological properties, clinical evidence, risks, and quality issues is clearly presented and well referenced. Phytoestrogen comes out rather well (reducing bone resorption in postmenopausal women) as does garlic, which at least in high dose reduces the frequency of tick bites. Red clover is safe but ineffective while shark cartilage is probably not even safe. A valuable series of tables summarizes herbs with adverse effects and interactions with antidiabetic and cardiac drugs, anticoagulants and oral contraceptives. Section 5 considers 38 conditions commonly treated with CAM (other conditions are indexed and discussed under individual therapies). The conclusions regarding atopic eczema—that conventional steroid treatments cannot be matched by CAM, and that therapies with the most promising evidence for efficacy are those with a psychological component—confirm my own impressions from the skin clinic. For migraine, biofeedback and possibly feverfew are effective and safer than conventional drug therapy. Similarly gingko is promising for tinnitus. Acupuncture and ginger are helpful for motion sickness and postoperative nausea. For many therapies there is insufficient evidence to recommend them, but this deficiency must not be confused with lack of effect. What about the wisdom of centuries of sages? Surely years of experience reflected in ‘expert opinion’ must be valuable? The authors actually tested this hypothesis, by comparing CAM recommendations with evidence from their systematic reviews. Not only was there lack of agreement between seven general CAM textbooks, but in some cases a therapy was recommended when there was conclusive trial evidence that it was ineffective or even contraindicated. The final section deals with general topics including legal, ethical and safety issues. Users of CAMs perceive them to be safer than conventional treatments, but must understand that ‘natural’ is not necessarily safe, unregulated preparations may be impure, and CAM providers sometimes delay or hinder access to potentially life-saving treatment if they cannot diagnose medical disorders. These concerns could all be addressed by better training and regulation of CAM providers. The bottom line is that CAM is amenable to rigorous testing and some forms of CAM are supported by evidence and do have a role in modern healthcare. The information in this book is balanced, authoritative and accessible. The enclosure of a CD-rom, with Medline links to all references as well as the whole text of the book, makes it astonishingly good value. The index might be improved in the next edition, but terms not listed (e.g. acne and psoriasis) can easily be found by searching the disc. This is a work which I will refer to every time a patient asks me about an alternative treatment.