Defined as the psychophysiologic response associated with placebos, the placebo effect is of considerable interest to researchers and clinicians. To ensure that study participants remain blinded to treatment, a placebo must resemble the investigational product in all aspects except for physiological activity: it should have the same shape, colour, delivery mode, smell and taste. To produce a placebo with all of these qualities, the development cost and, thus, the daily cost of providing them to patients in a clinical trial, can be significant. Cost-effectiveness analyses (CEAs) typically use efficacy and safety data from RCTs. In fact, phase 3 RCT data are often considered the most robust data source in CEAs. By subtracting the clinical effect in the placebo arm from the clinical effect in the active arm, CEAs remove the placebo response from the effectiveness side of the equation. However, the same method is not applied to the cost side: instead of subtracting from the cost of active treatment, the cost of placebo is ignored. This leads to an inaccurate estimation of the incremental cost of treatment relative to the incremental effects and, consequently, of the incremental cost-effectiveness ratio (ICER). We propose a method whereby both the costs and effects of placebo are incorporated into CEA. A CEA of Sativex in oncology pain will be used to illustrate the proposed approach. Results will be presented. In recognition of the clinical benefit that can be effected via the placebo response, RCTs have been designed to measure this response and to deduce the true effect of an active therapy. CEAs, which typically use data from these RCTs, should adopt a similar approach. Economic analyses should not only consider the effect but also the costs of placebo to achieve a more accurate prediction of the ICER for an active therapy.