Research · Explainer

What déjà vu reveals about how the brain processes memories

Déjà vu is a brief, unsettling sense that a brand-new moment has been lived before, paired with the knowledge that this cannot be true, and research suggests it can arise when a familiarity signal appears without conscious access to its source, creating a conflict with the knowledge that the situation is new. Because the person notices the clash between “this feels known” and “this is new”, déjà vu has become a useful window on how the brain judges whether something has been met before and how it checks those judgements.

This explainer sets out what déjà vu is and how often it occurs, the main explanations researchers have proposed, the experiments that have managed to trigger it in the laboratory and what is known about the brain regions involved. It also looks at how ordinary déjà vu differs from the kind that comes with epileptic seizures. The account draws on a landmark review of the phenomenon, a 2024 meta-analysis of 46 studies, virtual reality and brain-imaging experiments, and stimulation studies in people with epilepsy.

Woman pausing in a sunlit hallway with a puzzled expression, as if sensing she has been there before
A sudden sense of having stood in this exact spot before Editorial illustration

What is déjà vu?

The French phrase means “already seen”, but the experience is not really about sight. It can attach to a place, a conversation, a sound or a whole situation.

O’Connor and Moulin (2010) defined déjà vu as a feeling of recognising the present situation combined with awareness that the recognition is inappropriate. That second ingredient matters. Because the person knows the feeling is mistaken, déjà vu rarely changes what they do.

A related but different experience, sometimes called déjà vécu or “already lived”, involves a false sense of actually remembering the event rather than simply finding it familiar. Unlike ordinary déjà vu, it tends to come without insight that something is wrong, and it has been reported mainly in people with dementia, where it is often accompanied by confabulation (O’Connor and Moulin, 2010).

How common is it?

Déjà vu is one of the most widely shared odd experiences. In the review “A review of the déjà vu experience” (2003), Brown pulled together decades of survey data and concluded that about 60% of people report having had it. A later meta-analysis put the figure higher. Hadzic and Andersson (2024) pooled 10 studies of healthy people and estimated that 74% had experienced déjà vu not linked to any seizure.

Episodes are short and infrequent for most people. In one study included in the 2024 meta-analysis, 42.8% of people reported déjà vu a few times a year, 6.8% a few times a month and 0.9% weekly, while a further 22.5% said it happened less than once a year. About two-thirds of episodes lasted only a few seconds, and episode length was broadly similar whether déjà vu occurred in healthy people, between seizures or during a seizure (Hadzic and Andersson, 2024).

Much of this knowledge depends on [how researchers ask the question](/research/chinese-tourism-questionnaires-language-literacy-cultural-barriers-research). Brown (2003) noted that the literature was built from three kinds of evidence: retrospective surveys that ask people to recall past episodes, prospective surveys in which participants log episodes as they happen, and individual case studies. Each approach has weaknesses, and Brown concluded that more systematic research into both the prevalence and the causes of déjà vu was still needed.

Who gets it most?

Patterns in the survey data are consistent. Déjà vu becomes less frequent with age (Brown, 2003; Hadzic and Andersson, 2024). Brown (2003) also found that reports rose with education and socioeconomic level and appeared to be linked with stress and fatigue. In the 2024 meta-analysis, ordinary déjà vu showed no association with sex (Hadzic and Andersson, 2024).

Among people with epilepsy, the picture is less tidy. Déjà vu between seizures also became less common with age, but the single study that examined age and seizure-related déjà vu found no such link, and findings on sex differences for seizure-related episodes were inconsistent (Hadzic and Andersson, 2024).

What causes déjà vu?

No single cause has been agreed, but the candidate explanations fall into a few families. Brown (2003) grouped them into four:

  • dual processing, in which two mental processes that normally run together briefly fall out of step;
  • neurological explanations, such as a small seizure-like discharge or a hiccup in the timing of nerve signals;
  • memory explanations, in which something in the present is genuinely familiar from the past, but its source cannot be recalled;
  • attentional explanations, in which a scene is first taken in without attention and then perceived again with full attention, so the second look feels like a repeat.

Most laboratory research since then has tested the memory family, because it makes clear predictions that can be checked. The sections below describe the main lines of that work.

A familiar layout with a forgotten source

The best known experiments used virtual reality. In the study “Familiarity from the configuration of objects in 3-dimensional space and its relation to déjà vu: A virtual reality investigation” (2012), Cleary and colleagues showed students immersive 3D scenes through a headset. Later, they presented new scenes. Some of these had a completely different layout, while others copied the spatial arrangement of an earlier scene but used different objects, so a hedge garden might mirror the layout of a junkyard seen earlier.

When shown a copied layout, participants managed to recall the original scene only about 41% of the time, which left many cases in which the resemblance could be felt but its source stayed hidden. In those cases, the copied layouts still felt more familiar. In the first experiment, with 24 students, familiarity ratings for such scenes averaged 3.23 on a 0–10 scale, compared with 2.27 for unrelated layouts, and the probability of reporting déjà vu was 0.27 compared with 0.17 (Cleary et al., 2012).

A second experiment with 18 students confirmed the pattern and added a telling detail: scenes that participants had actually seen before but wrongly judged as new produced the most déjà vu of all, followed by new scenes with a copied layout and then by unrelated scenes (Cleary et al., 2012). The results support the idea that déjà vu can arise when the overall arrangement of a scene matches a memory that cannot be brought to mind.

Brief masked glimpses

The attentional family has also found laboratory support. Brown and Marsh (2009) showed participants unfamiliar symbols very briefly between visual masks and later found that these symbols were more likely to be judged as having been encountered before. The study measured this false sense of prior familiarity rather than déjà vu itself, but it shows how a fleeting prior glimpse can leave a trace that later feels like an earlier encounter. O’Connor and Moulin (2010) treated such induction studies with caution, noting that results can depend on how participants are asked about their experience.

Other laboratory routes, summarised in a 2010 review by O’Connor and Moulin, produced mixed results. In one method, participants first studied names of famous scenes and were later asked to identify pictures of famous scenes, some of which corresponded to the previously studied names. According to O’Connor and Moulin (2010), 33 of 38 participants, or 87%, reported at least one déjà vu experience during the procedure, although the reviewers cautioned that the high rate may have been influenced by how the question was asked. In a related method, déjà vu was reported in 17% and 26% of test trials but also in 13% and 23% of control trials, a gap small enough to be an artefact.

A study using hypnotic suggestions of familiarity produced déjà vu in 5 of 6 volunteers, although expectations created by the procedure may have played a part (O’Connor and Moulin, 2010). To make such studies easier to compare, O’Connor and Moulin (2010) recommended a clear working definition of déjà vu, questionnaires with several items instead of a single yes-or-no question, and room for participants to describe the experience in their own words.

A feeling of prediction that does not deliver

Many people say that during déjà vu they feel they know what will happen next. Cleary and Claxton tested that impression in the study “Déjà vu: an illusion of prediction” (2018).

Participants watched virtual tours that moved through a series of turns. Some tours copied the layout of an earlier tour that participants could not consciously recall. Just before the final turn, participants were asked which way the path would go.

Resemblance to an earlier tour increased reports of déjà vu, and déjà vu came with a stronger feeling of knowing where the path would turn. About half of the participants who reported déjà vu described a strong sense of premonition. Yet people in a déjà vu state were no better than chance at predicting the turn (Cleary and Claxton, 2018).

The authors concluded that the sense of premonition during déjà vu is a genuine feeling but an illusory one, produced by the déjà vu state itself rather than by any real knowledge of the future. They attributed it to a metacognitive bias, a distortion in how people judge their own knowledge, which may explain why déjà vu and feelings of premonition so often appear together.

What happens in the brain during déjà vu?

Evidence about the brain comes from two very different groups: people with epilepsy, whose seizures can produce déjà vu, and healthy volunteers in whom researchers try to create something similar. The two lines of evidence highlight different parts of the brain.

The temporal lobe and the rhinal cortex

Déjà vu can occur as a symptom of seizures involving the medial temporal lobe, an inner region of the temporal lobe involved in memory processing. In epilepsy surgery, electrodes are sometimes placed in this area to locate the source of seizures, and gentle electrical stimulation through them can produce experiences on demand. In a stimulation study of this kind, Bartolomei and colleagues (2004) found that stimulating the rhinal cortex, which lies next to the hippocampus and is thought to help produce feelings of familiarity, was more likely to trigger déjà vu than stimulating the hippocampus itself. Across studies pooled in the 2024 meta-analysis, stimulation-induced déjà vu was linked most often to the rhinal cortex, followed by the amygdala and hippocampus (Hadzic and Andersson, 2024).

Individual clinical cases point to a wider network. O’Connor and Moulin (2010) described a teenager whose déjà vu episodes, together with seizure warnings involving smell, stopped after surgical removal of the amygdala. In another patient, deep brain stimulation of the globus pallidus, a structure well outside the temporal lobe, evoked déjà vu along with increased blood flow in the right medial temporal region. The reviewers argued that such findings make sense only when brain activity is read in terms of connected networks rather than isolated regions.

A checking system in the frontal lobes

In healthy people, déjà vu may depend as much on the brain noticing an error as on the error itself. Urquhart and colleagues tested this in the study “fMRI evidence supporting the role of memory conflict in the déjà vu experience” (2021). They gave 21 volunteers lists of related words and used participants’ earlier letter-string counts as an additional novelty cue at test, allowing them to establish that a familiar critical lure had not actually appeared. Of the 21 volunteers, 16 reported déjà vu during the task.

The authors described the work as the first brain-imaging experiment to capture a laboratory version of déjà vu in healthy volunteers. Déjà vu reports were most numerous for related critical lures that participants correctly identified as new, although the difference from unresolved critical lures was not statistically significant (p = .147).

Brain scans showed that this conflict engaged regions linked to monitoring and decision-making, including the anterior cingulate cortex, medial prefrontal cortex and parietal cortex (Urquhart et al., 2021). The authors suggested that resolving memory conflict may be central to déjà vu. On this view, déjà vu in healthy people may be a sign that the brain’s checking system is working, catching a false familiarity signal before it is taken at face value.

How is everyday déjà vu different from déjà vu in epilepsy?

Not all déjà vu is the same. Researchers separate three kinds: non-ictal déjà vu in people without epilepsy, interictal déjà vu that people with epilepsy experience between seizures, and ictal déjà vu that forms part of a seizure.

In the study “Non-ictal, interictal and ictal déjà vu: a systematic review and meta-analysis” (2024), Hadzic and Andersson screened 1,677 records and included 46 studies from 13 countries. Their pooled estimates showed that déjà vu is common in all groups, but its character differs.

Ordinary episodes were mostly neutral or even pleasant, while seizure-related episodes were the most negative, with fear often reported. Feelings of derealisation and depersonalisation were more prominent in seizure-related déjà vu. However, derealisation was also associated with ordinary déjà vu, while findings on depersonalisation were inconsistent. General dissociation was not associated with ordinary déjà vu (Hadzic and Andersson, 2024).

Epileptiform activity has been recorded during seizure-related déjà vu, whereas limited EEG observations of ordinary déjà vu have not shown such activity. Neither ordinary nor seizure-related déjà vu predicted how well people performed on recognition memory tests, so frequent déjà vu does not appear to signal a poor memory (Hadzic and Andersson, 2024).

Type of déjà vu

Who experiences it

Pooled prevalence

Number of studies

Typical emotional tone

Non-ictal

People without epilepsy

74%

10

Mostly neutral or positive

Interictal

People with epilepsy, between seizures

62%

2

More negative

Ictal

People with epilepsy, during seizures

22%

11

Most negative, often with fear

Source: Hadzic and Andersson, 2024.

The authors argued that these forms should not be treated as one phenomenon, and proposed that ictal déjà vu be considered for listing as a symptom of epilepsy in future revisions of the international disease classification (Hadzic and Andersson, 2024). Seizure-related déjà vu was most common in familial mesial temporal lobe epilepsy and least common in occipital lobe epilepsy, which fits the link with the medial temporal lobe.

The meta-analysis has limits that its authors acknowledged. Reports of the subjective déjà vu experience relied on participants' descriptions, although some studies also used objective methods such as EEG recordings, brain imaging and electrical stimulation. Study quality varied, and case reports could not be pooled statistically. Differences between study designs produced considerable variation in the estimates, and only prevalence could be combined in numbers; other outcomes were summarised in words, often without figures (Hadzic and Andersson, 2024).

Déjà vu in other conditions

Déjà vu has also been studied in psychiatric conditions. Reviewing this work, O’Connor and Moulin (2010) noted that people with schizophrenia reported déjà vu less often than comparison groups but found it more distressing. In people with anxiety disorders, between a third and a half were reported to experience déjà vu or déjà vécu. These findings suggest that how déjà vu feels, and not only how often it occurs, can differ between groups.

What is proven about déjà vu, and what is not

Several points are now well supported by the studies described above:

  • Most people experience déjà vu at some point, with estimates of 60–74% in healthy populations (Brown, 2003; Hadzic and Andersson, 2024).
  • It becomes less frequent with age and usually lasts only seconds (Hadzic and Andersson, 2024).
  • Scenes whose layout matches a forgotten scene can trigger déjà vu in the laboratory (Cleary et al., 2012).
  • The feeling of knowing what comes next during déjà vu does not improve real predictions (Cleary and Claxton, 2018).
  • Stimulating the rhinal cortex can trigger déjà vu, and seizure-related déjà vu differs from ordinary déjà vu in tone and brain activity (Bartolomei et al., 2004; Hadzic and Andersson, 2024).

One leading explanation suggests that déjà vu arises when a sense of familiarity conflicts with the knowledge that a situation is new. Laboratory findings support the involvement of memory-conflict monitoring, although the mechanisms behind spontaneous everyday déjà vu remain uncertain.

Several questions remain open. All evidence on everyday déjà vu relies on [people’s own reports](/research/accuracy-of-self-reported-participation-rates-research), which are hard to verify, and laboratory versions may not match the spontaneous experience, a problem long recognised in research [methodology](/research?sub=methodology).

The four families of explanation are not mutually exclusive, and it is possible that different routes produce similar feelings. The estimate for déjà vu between seizures rests on only two studies. Why déjà vu fades with age is unclear, and so is the reason it seems more common in people with more education.

For readers drawn to [scientific curiosities](/research?sub=curiosities), déjà vu shows how an odd everyday sensation can be turned into a testable research question. Summaries of [what individual studies found](/research?sub=what-the-study-found) give more examples of experiments that turned a familiar experience into measurable data.

Frequently asked questions

Is déjà vu a sign of a health problem?

For most people, no. Déjà vu is common among healthy people and is usually brief (Hadzic and Andersson, 2024). It can also occur during epileptic seizures, sometimes alongside fear or other symptoms. Researchers distinguish these seizure-related episodes from ordinary déjà vu because they can involve different experiences and brain activity.

Why does déjà vu happen less with age?

Survey studies consistently find that déjà vu becomes less frequent in older adults (Brown, 2003; Hadzic and Andersson, 2024). The reason is not known, because the studies that report the decline were designed to measure how often déjà vu occurs rather than to explain why its frequency changes over a lifetime.

Can déjà vu predict the future?

No evidence supports that idea. In a virtual reality study, people in a déjà vu state felt more confident about which way a path would turn next, yet their predictions were no better than chance (Cleary and Claxton, 2018). The sense of foreknowledge appears to be a side effect of the familiarity feeling itself.

What is the difference between déjà vu and déjà vécu?

Déjà vu is a feeling of familiarity combined with the knowledge that it is mistaken. Déjà vécu, or “already lived”, involves a false sense of actually remembering the moment, usually without awareness that anything is wrong. It has been described mainly in people with dementia and is often accompanied by confabulation (O’Connor and Moulin, 2010).

Sources

  1. Brown, A.S. A review of the déjà vu experience. Psychological Bulletin, 2003, 129(3), 394–413. DOI 10.1037/0033-2909.129.3.394
  2. O’Connor, A.R., Moulin, C.J.A. Recognition without identification, erroneous familiarity, and déjà vu. Current Psychiatry Reports, 2010, 12(3), 165–173. DOI 10.1007/s11920-010-0119-5
  3. Hadzic, A., Andersson, S. Non-ictal, interictal and ictal déjà vu: a systematic review and meta-analysis. Frontiers in Neurology, 2024, 15, 1406889. DOI 10.3389/fneur.2024.1406889
  4. Cleary, A.M., Brown, A.S., Sawyer, B.D., Nomi, J.S., Ajoku, A.C., Ryals, A.J. Familiarity from the configuration of objects in 3-dimensional space and its relation to déjà vu: a virtual reality investigation. Consciousness and Cognition, 2012, 21(2), 969–975. DOI 10.1016/j.concog.2011.12.010
  5. Cleary, A.M., Claxton, A.B. Déjà vu: an illusion of prediction. Psychological Science, 2018, 29(4), 635–644. DOI 10.1177/0956797617743018
  6. Brown, A.S., Marsh, E.J. Creating illusions of past encounter through brief exposure. Psychological Science, 2009, 20(5), 534–538. DOI 10.1111/j.1467-9280.2009.02337.x
  7. Bartolomei, F., Barbeau, E., Gavaret, M., Guye, M., McGonigal, A., Régis, J., Chauvel, P. Cortical stimulation study of the role of rhinal cortex in déjà vu and reminiscence of memories. Neurology, 2004, 63(5), 858–864. DOI 10.1212/01.WNL.0000137037.56916.3F
  8. Urquhart, J.A., Sivakumaran, M.H., Macfarlane, J.A., O’Connor, A.R. fMRI evidence supporting the role of memory conflict in the déjà vu experience. Memory, 2021, 29(7), 921–932. DOI 10.1080/09658211.2018.1524496
Marcus Reed

Marcus Reed — author

Marcus Reed writes with an editorial focus on statistics and quantitative research, including effect sizes, regression models, uncertainty, associations and the numerical results reported in scientific studies. At Cabdirect, Marcus covers quantitative findings...

Related