SummaryIn addition to the actual pharmacological effect of a drug, its efficacy depends on a number of non-specific factors globally termed the placebo effect. To distinguish the pharmacological effect from the non-specific effect it may be justified to compare a supposedly active drug to a placebo, a “pseudo-drug” identical in appearance but pharmacologically inert.Factors underlying the placebo effect include the patient’s personality (placebo reactor or non-reactor), the convictions and enthusiasm of the medical staff, the conditions of administration and the appearance of the drug.The placebo effect does not appear to present distinctive characteristics which distinguish it from a true pharmacological effect and it may even be associated with adverse effects, known as a nocebo effect.A placebo may be used in comparison with a pharmacologically active preparation or it may form a three-way comparison with two supposedly active preparations to ensure that the criterion for assessment is sufficiently sensitive. Placebos can also be used for a double-blind comparison of two drugs that cannot be made to look alike (double-dummy technique), to compare drug combinations with their individual constituents, and finally during run-in, withdrawal or wash-out periods.Selecting placebo non-reactors for a therapeutic trial is sometimes possible to make the comparison more sensitive.It is extremely important that the placebo match the active preparation as closely as possible.So-called “impure placebos” (i.e. drugs of dubious efficacy or preparations producing the same side-effects as the active drug under study) should be avoided since the results may subsequently prove impossible to interpret.Finally, the ethics of using a placebo should be given due consideration, if possible by an independent committee.