The place of placebos in surgical trials is controversial, with little published guidance on when they should be used, other than that published by the American Medical Association in 2002. Our research group recently completed a 24 month in-depth investigation of the feasibility and acceptability of conducting a full-scale UK multicentre placebo-controlled trial of surgical arthroscopic lavage for the management of knee osteo-arthritis (the KORAL study). This involved: focus groups, interviews and surveys with key stakeholder groups (surgeons, anaesthetists, prospective participants and members of research ethics committees) to identify the most appropriate form the placebo should take and to ascertain the acceptability of a placebo-controlled design within the surgical context; and a formal pilot in which patients were recruited to a trial with a placebo control arm. The findings suggested that the optimal circumstances for a placebo-controlled trial in surgery are those in which: a) alternative designs would provide inferior (and potentially biased) results, particularly where the primary outcome is of a subjective nature; b) a placebo procedure can be devised which can adequately mimic the active intervention with an acceptable level of intrusiveness and risk to potential participants; c) potential participants can judge for themselves whether the risk-to-benefit ratio of participation is acceptable to them (after assessing clear descriptions and careful explanations of the advantages and disadvantages of taking part); and d) appropriate practical arrangements can be instituted in local centres to ensure that the delivery of such a design would be feasible. These findings and their implications for the field of surgical trials will be discussed.