One of the problems in assessing acupuncture efficacy has been the lack of a standardised placebo/control. Despite the lack of 'proof', acupuncture is widely used; hence it would seem that a large proportion of the public are not too concerned with the question of efficacy. Patients and researchers approach the issue of whether acupuncture "works" in an entirely different way. It may be that the management of chronic pain as a whole is largely context driven and non-specific. Therefore, research should focus on areas such as pragmatic/comparative trials and the non-specific effects of treatment. Improving our understanding of the neural mechanisms and substrates of acupuncture, placebo and non-specific effects might enable us to better define a "true placebo" and improve trial design. Imaging studies, however, need to be much more pragmatic. Because of the large overlap in areas of brain activated through acupuncture, pain, placebo and non-specific factors, separating these out in an attempt to pinpoint the mechanisms behind acupuncture will be difficult. Ultimately we need a balance between efficacy, comparative and mechanistic trials using imaging work to inform the whole picture. A broader view of research is therefore necessary to yield meaningful answers and we need to look at the whole package that acupuncture delivers.