To examine whether written informed consent might influence the results of clinical trials the effect of placebo when given with or without informed consent to patients suffering from insomnia was studied. The study was a single blind observer blinded trial, and patients were paired according to sex, age, and hospital environment. Randomisation assigned the first patient of each pair to the control group (without informed consent) or the group to give informed consent. Of the 56 patients, 26 refused to give informed consent, and the age and sex distribution of these differed significantly (p<002) from the 30 pairs of patients ultimately enrolled into the study. In this "biased" sample, the hypnotic activity of placebo was significantly higher in the control group (p<005).It is concluded that the informed consent procedure biases the results of clinical trials and might affect their general applicability.