We evaluated the influence of hypnotizability, pain expectation, placebo analgesia in waking and hypnosis on tonic pain relief. We also investigated how placebo analgesia affects somatic responses (eye blink) and N100 and P200 waves of event-related potentials (ERPs) elicited by auditory startle probes. Although expectation plays an important role in placebo and hypnotic analgesia, the neural mechanisms underlying these treatments are still poorly understood. We used the cold cup test (CCT) to induce tonic pain in 53 healthy women. Placebo analgesia was initially produced by manipulation, in which the intensity of pain induced by the CCT was surreptitiously reduced after the administration of a sham analgesic cream. Participants were then tested in waking and hypnosis under three treatments: (1) resting (Baseline); (2) CCT-alone (Pain); and (3) CCT plus placebo cream for pain relief (Placebo). For each painful treatment, we assessed pain and distress ratings, eye blink responses, N100 and P200 amplitudes. We used LORETA analysis of N100 and P200 waves, as elicited by auditory startle, to identify cortical regions sensitive to pain reduction through placebo and hypnotic analgesia. Higher pain expectation was associated with higher pain reductions. In highly hypnotizable participants placebo treatment produced significant reductions of pain and distress perception in both waking and hypnosis condition. P200 wave, during placebo analgesia, was larger in the frontal left hemisphere while placebo analgesia, during hypnosis, involved the activity of the left hemisphere including the occipital region. These findings demonstrate that hypnosis and placebo analgesia are different processes of top-down regulation. Pain reduction was associated with larger EMG startle amplitudes, N100 and P200 responses, and enhanced activity within the frontal, parietal, and anterior and posterior cingulate gyres. LORETA results showed that placebo analgesia modulated pain-responsive areas known to reflect the ongoing pain experience.