Previous research by Stam and Spanos suggests that if waking analgesia is followed by hypnotic analgesia, subjects refrain from maximally responding during the waking trial so they report less pain under hypnosis (i.e., a "holdback effect"). This hypothesis was re-examined using more stringent controls. Thirty-six highly susceptible subjects chosen by a combination of the Harvard Group Scale of Hypnotic Susceptibility, Form A and the Stanford Hypnotic Susceptibility Scale, Form C were randomly assigned to one of three treatment groups (waking analgesia followed by hypnotic analgesia, waking analgesia followed by waking analgesia, or hypnotic analgesia followed by waking analgesia). Each group received three 60-second immersions of cold pressor pain stimulation (baseline, Immersion 1, Immersion 2) and rated pain using a magnitude estimation and a category rating scale. The obtained results failed to support the hypotheses of a holdback effect or a "reverse-order holdback effect." Properties of within-subjects and between-subjects designs were considered in explaining the superiority of hypnotic analgesia over waking analgesia typically found in within-subjects models.