If psychotherapy intended to achieve a status of scientific recognition and a health policy validation, must do so through a convincing empirical support (EST: Empirically Supported Treatments). This stance, if well seemingly sensible, finds several difficulties: one is the difference between the searchable scientifically on one side, and clinical practice most preferences and values of the patient by another; This is intended to save the furniture through a broader redefinition that it integrates everything (EBP: Evidence-Based Practice). However, I would like to focus a very specific aspect of scientific research of psychological therapies: how to validate them empirically (the Task Force of the APA requires a psychological treatment validated empirically at least two studies showing superiority over placebo or other bona fide treatment). But, as well as a drug should demonstrate superiority over placebo in double-blind design (and supposedly on a control group on waiting list), how do this in psychotherapy? The use of placebos is problematic even in pharmacological studies, since many patients can guess which group they are (unblinding) from the presence/absence of side effects (which weakens the placebo group and power placebicamente to the active group, potentially distorting the results in favor of the drug), but at least have the conceptual clarity of be trying to control the psychological effects of a particular treatment for this both conditions must share their psychological properties to maximize: color, form, route of administration, information, etc. Now, what would be a placebo in psychotherapy and how to incorporate it in a double-blind? Input, there is a conceptual impediment: psychotherapy is based on psychological change factors (know or not are), but this is true also for placebo: both play with meanings (although some psychotherapies are supposed to include other mechanisms additional; You can than some Pavlovian placebo also); If we asked the maximum psychological equivalence, placebo would be the treatment. And if we find a placebo with no psychological effect mode? Well, then no longer be a control. On the other hand, there is a practical impossibility: should be designing convincing placebo psychotherapies that patient bought as if they were psychotherapeutic treatments but that they were not at all (complicated) and also make out blind Administrator status (there already directly impossible).