The placebo response in medicine and psychology is incompletely understood, and is usually either mystified or down-played; a rational approach to the understanding of the placebo response, however, can be deduced from the scientific literature. We present 3 theories that may explain most of the placebo response in medical and psychological interventions: a) Regression to the mean (RTM): This refers to "errors" in measuring the outcome of therapies, that are due to small samples sizes, global assessment variables, spontaneous variability of symptoms, and other methodological reasons. b) Pavlovian Conditioning (PC): here the placebo response occurs as the consequence of successful association of diagnostic and therapeutic procedures in the past (illness history) of the individuum that resulted in symptom improvement. Neutral stimuli in the context of a treatment can thus gain therapeutic potency (become conditioned stimuli), e. g. the procedure of an injection or the colour of a drug. c) Signal detection theory (SDT): manipulation of expectencies and cognitions of the patient by suggestions and verbal instructions of the health care system will change the willingness of the patient to perceive symptoms as improved/worsened, specifically if this happens under "noisy" circumstances, e. g. with high spontaneous variability of symptoms. All three models are illustrated with examples from medical and psychological treatments or experiments, including recent findings of cortical correlates of the placebo response in functional brain imaging investigations. Potential biological mechanisms for the placebo response are discussed, including the possibility of genetic predisposition to be a placebo responder.