Hypnotic suggestion is an empirically validated form of pain control but it remains uncertain whether the use of a formal hypnotic induction is a necessary component. The precise underlying mechanism of pain control with suggestion also remains unclear. Following previous studies, we hypothesized that suggested pain control would occur via direct alteration of activity in pain-related neural centres in both patients altering clinical pain and healthy volunteers altering experimentally induced heat pain (1,2).Patients with fibromyalgia and healthy controls were pre-selected for high hypnotic suggestibility, and for the ability to respond to suggestions for pain control. Patients received suggestions to increase and decrease their clinical pain. Controls received a noxious heat pain experience and received identical suggestions to increase and decrease their pain. Suggestions were delivered before and after a hypnotic induction and BOLD activity was measured concurrently. Both hypnotic and non-hypnotic suggestion produced large and highly significant increases and decreases in clinical and experimental pain report (see figure 1). Significantly larger changes in report followed hypnotic suggestion for both groups, though these differences were relatively small in absolute terms. Thus, the behavioral results overall indicated small differences between the effects of hypnotic and non-hypnotic suggestion and no differences when altering clinical versus experimental pain. BOLD responses were measured with fMRI at the same time that the suggestions were delivered. A series of contrasts were performed to compare medium (baseline) with low and high pain conditions. Separate contrasts were conducted for suggestions with and without hypnosis, and for the patients and controls. In the patients, significantly increased BOLD responses were observed in the thalamus, anterior cingulate cortex, insula and primary and secondary somatosensory cortices. These changes were greater for hypnotic compared with non-hypnotic suggestion. In contrast, only significantly decreased BOLD activity was observed in control subjects, largely in the thalamus, anterior and posterior cingulate cortices. These changes were greater for hypnotic compared with non-hypnotic suggestion. To further explore this pattern of response, ROI values were calculated and plotted for a series of regions typically associated with pain. The pattern of response in the patients was generally linear with increasing pain report, especially following hypnotic suggestion (see figure 2). The pattern of response in the controls was very different and was generally higher when there was a suggestion to increase or decrease pain from the baseline, which was also especially evident following hypnotic suggestion. In conclusion, the behavioral reports of altered pain experience revealed only small differences between hypnotic and non-hypnotic suggestions and no differences across clinical and experimentally-induced pain. Based on behavioral report alone, the mechanism of suggestion could be interpreted as largely the same regardless of the presence or absence of a hypnotic induction or the type of pain experience. The fMRI data, however, indicate a very different interpretation. First, the hypnotic induction produces larger changes in brain activity and second the pattern of brain activity is radically different for clinical compared with experimental pain. These findings imply that the induction has an important effect on the underlying neural areas mediating the effects of suggestion. Further, they imply that the mechanism of suggestion in patients altering clinical pain is different from the mechanism altering experimental pain in controls. The simplest interpretation of the patient response is that the suggestions altered pain experience via corresponding changes in pain-related brain regions, whereas in controls suggested changes in pain experience engaged cognitive control, which resulted in decreased BOLD response when comparing medium (baseline) with low pain.