Forty-two individuals selected for high hypnotizability or for low hypnotizability were taught lists of words during hypnosis and assessed for recognition following hypnosis using event-related potential (ERP) procedures, both before and after the cue to reverse amnesia. A subgroup of low-hypnotizable participants were asked to simulate hypnotic behavior. All participants had larger late positive component (LPC) amplitudes to learned than to unlearned words, regardless of whether amnesia was reported. The highly hypnotizable participants who reported recognition amnesia, however, had significant changes in attention-related (P1 and N1) and recognition-related (N400 and LPC) ERP component amplitudes as a function of whether amnesia was reported. These data suggest that posthypnotic amnesia may involve alterations in the processes of attention, selection, and accessibility.