Overview

The formal scientific study of déjà vu is fairly young, even though the experience itself is ancient. The term entered the scientific literature in the late 1800s. French philosopher Émile Boirac coined it, though descriptions of the feeling go back to antiquity.^[8] The most widely used definition was written by neuropsychiatrist Vernon Neppe in his 1979 doctoral thesis. He called déjà vu “any subjectively inappropriate impression of familiarity of the present experience with an undefined past.”^[5] Every word matters. The familiarity is “subjectively inappropriate” because the person knows it cannot be right. It concerns a present experience, not a memory. And the “past” is “undefined” because no specific prior event can be identified.

What makes déjà vu both interesting and hard to study is that it is ordinary and strange at the same time. The brain sends a strong familiarity signal. At the same moment, your rational mind insists no such prior encounter ever happened. Neuroscientist Akira O’Connor of the University of St Andrews describes it as “a conflict between the sensation of familiarity and the awareness that the familiarity is incorrect.”^[9] He suggests it may actually be a sign of healthy brain function. The frontal regions of the brain detect the false familiarity signal and flag it as wrong. So what feels like a glitch may be the brain’s fact-checking system working correctly.

On the brain side, evidence points to a network of mesiotemporal regions. The hippocampus and entorhinal cortex appear central to producing déjà vu in both epileptic and non-epileptic people.^[3] Studies of temporal lobe epilepsy (TLE) patients have been especially useful. In TLE, déjà vu can appear as a warning sign just before a seizure. Researchers can deliberately trigger the experience by electrically stimulating the rhinal cortices. EEG recordings show that synchronized firing between the rhinal cortices and the hippocampus or amygdala increases during these triggered episodes.^[10] The brain chemical dopamine is also involved. Dopamine plays a role in signaling familiarity, and drugs that affect dopamine levels are linked to more frequent déjà vu reports.^[11]

Despite this progress, déjà vu is notoriously hard to study in a lab. It happens spontaneously and lasts only seconds.^[12] This has pushed researchers to get creative, most notably by using virtual reality to reliably trigger déjà-vu-like states under controlled conditions.

Research landscape and major figures

Modern experimental déjà vu research has been shaped by two overlapping traditions: cognitive memory research and parapsychological phenomenology (the careful description of unusual experiences).

Cognitive and neuroscientific tradition. Psychologist Alan Brown at Southern Methodist University produced the most thorough early overview of the field. His 2003 review in Psychological Bulletin and his 2004 book The Déjà Vu Experience catalogued epidemiological data and theoretical explanations. He found that déjà vu is typically tied to an entire setting rather than to specific objects, sounds, or people. It is also unrelated to gender or race.^[2] Brown counted more than 50 distinct explanations that had been proposed. That number later grew to over 70. They fall broadly into perceptual, memory-based, and physiological categories.^[7]

Anne Cleary at Colorado State University pioneered the use of virtual reality to bring déjà vu into the lab. In a landmark 2012 study, Cleary and colleagues built immersive VR scenes. Some new scenes had the same spatial layout as scenes participants had viewed earlier but could not consciously recall. Those scenes produced significantly higher familiarity ratings and more déjà vu reports than completely novel scenes did.^[4] This supported what is called the Gestalt familiarity hypothesis. The idea is that déjà vu happens when the overall arrangement of elements in a current scene matches a stored memory the person cannot consciously access.^[13] Crucially, participants still recognized the scene as new even while feeling it as familiar. That combination had never been reproduced in a lab before.^[14]

Cleary’s 2018 work with Alexander Claxton, published in Psychological Science, tackled the premonition side of déjà vu. Many people report that during the experience they feel they know “what will happen next.” The experiments showed that déjà vu did not give people any real ability to predict the next turn in a navigational path. Their actual prediction accuracy was no better than chance. But the feeling of knowing was real as an experience.^[6] A 2025 study by Huebert, Myers, and Cleary in Consciousness and Cognition went further. It provided experimental evidence that familiarity detection itself causes the illusory feeling of being able to predict the future. This happened even when participants were not prompted to think about déjà vu at all.^[15]

Phenomenological and parapsychological tradition. Vernon Neppe is a neuropsychiatrist and former president of the South African Society for Psychical Research. He has made the most sustained effort to connect neuroscience and parapsychology in déjà vu research. Starting with his PhD thesis in 1981 and codified in his 1983 book The Psychology of Déjà Vu: Have I Been Here Before?, Neppe used a statistical technique called multidimensional scaling across 22 dimensions. He found four distinct subtypes of déjà vu, each linked to a different population and a different cause:^[5]

• Associative déjà vu — the most common type, found in healthy people without epilepsy, psychosis, or reported psi experiences. Considered non-pathological.
• Temporal lobe epileptic (TLE) déjà vu — occurs as a pre-seizure warning in patients with temporal lobe epilepsy, with distinct neurological features.
• Psychotic déjà vu — occurs in people with schizophrenia and has a qualitatively different character.
• Subjective paranormal experience (SPE) déjà vu — reported by people who also report other psi-type experiences. It is described as more intense, longer-lasting, and accompanied by a specific “predictive knowing” quality and time distortions.

The four groups separated clearly in multivariate analysis. This supports the view that “déjà vu” is not one phenomenon but a family of distinct experiences with different causes.^[16] Neppe’s work also built on earlier parapsychological thinking by C. T. K. Chari. In 1964, Chari proposed a three-way division of déjà vu into pathological, normal, and precognitive/telepathic categories.^[17]

The parapsychological angle

Parapsychology connects to déjà vu mainly through three related topics: precognition research, reincarnation, and the study of unusual subjective experiences.

The most common parapsychological claim is that déjà vu represents a genuine flash of precognition. The idea is that the sense of “having been here before” reflects real foreknowledge from a precognitive dream or vision. Researchers have looked at this carefully. Factor analyses of déjà vu questionnaire data found that déjà vu clusters with precognitive dreams and memory-related factors. It does not cluster with dissociative experiences like depersonalization. This fits the pattern of people attributing their déjà vu to a “suddenly remembered” precognitive dream.^[18] Some parapsychologists have noted that a déjà vu episode can easily be confused with a precognitive dream that was later forgotten.^[19] However, Cleary’s VR work directly tested the predictive claim. It found that the feeling of knowing what comes next is illusory. Déjà vu does not give people actual predictive accuracy.^[6]

Neppe’s SPE subtype is the most carefully developed parapsychological hypothesis about déjà vu. This variety is reported by people who also report other psi-type experiences, including possible telepathy and precognitive dreaming. It is described as more intense and longer-lasting than ordinary déjà vu. It comes with a specific predictive “knowing” quality and distortions of time, both past and future.^[20] In some documented cases, people experiencing SPE déjà vu while visiting an unfamiliar location reportedly predicted what would be found in the next room.^[19] Neppe and his collaborators deliberately use the term “subjective paranormal experience” rather than “paranormal.” This keeps the claim neutral. The point is that the person experiences the episode as paranormal, not that it has been verified as such.^[16]

A separate parapsychological thread connects déjà vu to reincarnation cases research and past-life memory. This is especially true for a variant called déjà vécu, meaning “already lived through.” In déjà vécu, a person does not just recognize a scene. They feel they have fully lived through an entire prior sequence of events there. This connection is largely anecdotal and speculative in the peer-reviewed literature, though it appears in clinical case reports. Some commentators have also linked strong déjà vu to out-of-body experiences and other altered states, given shared features of dissociation and a sense of time being out of joint.

Déjà vu also creates a practical problem for precognition researchers. A déjà vu episode can be reframed after the fact as a confirmed precognitive dream. This inflates apparent hit rates and makes it harder to assess genuine psi claims. Some researchers have flagged it as a potential confound in experimental psi studies involving dreams and altered states.^[21]

Overall, déjà vu holds an instructive place in parapsychology. It is a real, measurable, near-universal human experience. The standard cognitive explanation, a mismatch in the brain’s familiarity system, accounts for most reported cases. But a small, phenomenologically distinct subset, Neppe’s SPE type, remains of active interest to researchers open to anomalous interpretations. Whether that subset turns out to be an unusually intense form of ordinary memory misfiring, or something more, is still an open empirical question.