Fractal analysis was applied to study the trends of EEG signals in the hypnotic condition. The subjects were 19 psychiatric outpatients. Hypnotizability was measured with the Hypnotic Induction Profile (HIP). Fifty-four sets of EEG data were analyzed by detrended fluctuation analysis (DFA), a well-established fractal analysis technique. The scaling exponents, which are the results of fractal analysis, are reduced toward white noise during the hypnotic condition, which differentiates the hypnotic condition from the waking condition. Further, the decrease in the scaling exponents during hypnosis was solely associated with the eye-roll sign within specific cortical areas (F3, C4, and O1/2) closely related to eye movements and attention. In conclusion, the present study has found that the application of the fractal analysis technique can demonstrate the electrophysiological correlations with hypnotic influence on cerebral activity.