The mediator function of response expectancies and the moderator function of hypnotic and non-hypnotic imaginative suggestibility were evaluated in the analogue treatment of pain. One hundred and sixty-seven participants previously assessed for hypnotic and imaginative suggestibility were randomly assigned to distraction, cognitive-behavioural package (i.e., Stress Inoculation Training), hypnotic cognitive-behavioural package, hypnotic analgesia suggestion, placebo control, or no-treatment control conditions. The four 'active' treatments reduced pain more than the no-treatment control condition. There was no statistical difference in effectiveness between these four treatments, but only the cognitive-behavioural package reduced pain more than the placebo control condition. Response expectancies partially mediated the effects of treatment on pain. Imaginative suggestibility, defined as a generalized tendency to respond to imaginative suggestions delivered outside of hypnosis, moderated the effects of the cognitive-behavioural package. Contrary to prediction, neither hypnotic suggestibility, nor hypnotizability (operationalized as hypnotic suggestibility with imaginative suggestibility statistically controlled) moderated the effects of the hypnotic treatments.