In 1968, a U.S. military–affiliated lab claimed to study how long people could face their own reflection. The truth, hidden for decades, was a controlled isolation experiment in an endless mirrored room that none escaped unchanged.
The public version of the story is dull.
In 1968, psychologists at a U.S. military-affiliated research lab claimed to run a “self-perception endurance test,” inspired by contemporary mirror-gazing research. The stated goal was to measure how long a person could tolerate constant visual contact with their own reflection.
But the classified reports — those never meant for public release — paint a much darker picture.
The Rooms
Fifteen volunteers were split into three test groups. Each subject was confined to an identical “observation cell”: no windows, no clocks, no human contact. The walls, floor, and ceiling were covered entirely in mirrors, and the space was kept under bright, constant illumination, eliminating all shadows.
Food was minimal — a nutrient paste served once a day through a slot in the door. Waste was removed in the same manner.
The groups were separated by exposure length:
- Group A: 15 days
- Group B: 30 days
- Group C: 60 days
Although durations varied, all participants passed through similar stages:
- Days 1–4: Irritability, pacing, and insomnia.
- Days 5–15: Heightened aggression — shouting at mirrors, striking the walls, attempting to damage reflective panels.
- Days 15+: Rapid physical decline, refusal of food, and periods of unresponsiveness.
Severe weight loss was recorded across all groups. Starvation occurred most rapidly in Groups B and C. One Group B subject collapsed on day 19, weighing just 39 kilograms, and was removed. By the time others were withdrawn, skeletal protrusion and muscle atrophy were visible. Despite this, the experiment continued.
Observation Period: Day 1–15
By day 10, participants across all groups were showing signs of paranoia. One subject refused food entirely, claiming her reflection was “chewing for her.” Others described subtle inconsistencies in their reflections’ timing or expressions.
Reports of “reflection irregularities” began appearing consistently after 6–8 days of confinement. Subjects frequently described delayed mirroring or expressions they had not consciously made. These discrepancies were never observed by researchers, leading to the conclusion they were perceptual distortions caused by stress and prolonged self-observation.
Outcome for Group A:
When released on day 15, all members of Group A showed severe psychological distress and refused to enter rooms with reflective surfaces. They were prescribed antidepressants and placed in long-term therapy.
Unofficial notes describe “persistent fear response to self-image; aversion to reflective stimuli; perceptual distortion under controlled lighting conditions.”
Observation Period: Day 15–30
By day 18, two subjects requested complete darkness, stating that “someone else” was in the room.
On day 21, one participant’s reflection appeared to move with a one-to-two second delay — a phenomenon captured on video and confirmed by observers. The cause remains undetermined, though analysts suggested extreme fatigue and neurological dissociation may have contributed.
Day 24 brought the first major self-harm incident: a subject attempting to claw their own face, repeating, “It’s not me.” They were sedated and removed.
The remaining participants experienced extreme weight loss. One collapsed before day 30 and was carried out unconscious.
Outcome for Group B:
All surviving members were diagnosed with acute dissociative psychosis and transferred to long-term psychiatric facilities. None regained independent living capacity.
Observation Period: Day 30–60
By day 35, participants no longer responded to the intercom. Two began repeating short, meaningless phrases while facing the same mirrored corner for hours.
By day 44, multiple subjects claimed to be “repeating” the same 12-hour period. While this could not be verified, recordings showed occasional mismatches between subjects’ movements and their reflections’ timing, similar to the verified delay in Group B.
On day 51, two participants began a determined effort to obstruct the mirrors by spitting on surfaces and smearing nutrient paste across the glass. The attempt failed; the paste and moisture quickly slid away, either due to the mirror coating repelling liquid or the lighting keeping reflections visible despite smears.
By day 58, two subjects developed what the researchers termed “cross-identity dissociation.” Each displayed a secondary personality, but these identities appeared to be exchanged: one subject’s secondary state closely mirrored the other’s primary personality, and vice versa. Both secondary states produced similar phonetic vocalizations consistent with glossolalia or stress-induced language distortion, with analysis confirming no match to any known language. Communication in these states was minimal but notably synchronized in rhythm and tone.
Outcome for Group C:
No member of Group C completed the full 60 days in stable condition. All were placed under indefinite psychiatric care orders and removed from public record.
Aftermath and Erasure
The experiment was officially “terminated” in late 1968. Public statements claimed it had lasted only three days, halted after “unexpected stress reactions” in the volunteers.
Classified memos confirm the mirrored cells were dismantled under direct military order, with most panels destroyed and one possibly relocated to an unmarked government storage facility.
The “three-day” narrative was deliberately reinforced. Group A, having lost all perception of time, believed they had been inside only a few days, making the cover story difficult to challenge.
All ten participants from Groups B and C were either committed to psychiatric institutions under altered identities or removed from public record entirely. None returned to independent civilian life. Six died by suicide within five years of the experiment’s conclusion.
