Response expectancies have been proposed as the major determinant of placebo effects. Here we report that different expectations produce different analgesic effects which in turn can be harnessed in clinical practice. Thoracotomized patients were treated with buprenorphine on request for 3 consecutive days, together with a basal intravenous infusion of saline solution. However, the symbolic meaning of this basal infusion was changed in three different groups of patients. The first group was told nothing about any analgesic effect (natural history). The second group was told that the basal infusion was either a powerful painkiller or a placebo (classic double-blind administration). The third group was told that the basal infusion was a potent painkiller (deceptive administration). Therefore, whereas the analgesic treatment was exactly the same in the three groups, the verbal instructions about the basal infusion differed. The placebo effect of the saline basal infusion was measured by recording the doses of buprenorphine requested over the three-days treatment. We found that the double-blind group showed a reduction of buprenorphine requests compared to the natural history group. However, this reduction was even larger in the deceptive administration group. Overall, after 3 days of placebo infusion, the first group received 11.55 mg of buprenorphine, the second group 9.15 mg, and the third group 7.65 mg. Despite these dose differences, analgesia was the same in the three groups. These results indicate that different verbal instructions about certain and uncertain expectations of analgesia produce different placebo analgesic effects, which in turn trigger a dramatic change of behaviour leading to a significant reduction of opioid intake.