Hypnotic elicitation of multiple ego states was explored using Hilgard's "hidden observer" paradigm. Twenty subjects in two groups: hypnosis and simulation were utilized to examine the impact of experimental demand characteristics on the production of multiple ego states. Self-report and hypnotist-report measures were obtained in a test-retest design. Multiple t-tests and chi-square analyses were computed with significant differences on key multiple ego state items found between groups. Results demonstrated retest reliability, but not inter-rater reliability of this dissociative phenomena, since hypnotists failed to discriminate real from "faked" hypnotic involvement. Exploration of multiple ego states using non-hypnotic control conditions and multiple dependent measures is suggested for future research.