Death in Humans: An Analytical Report

Nov 12, 2026 · Brian Yegge

Death is a process that unfolds over minutes to hours, not an instant, and the brain can stay electrically active well into it. About one in five people revived from cardiac arrest report a vivid experience; most of those are peaceful, a minority are terrifying, and roughly four in five people recall nothing. No controlled study has yet shown a mind perceiving the world without a working brain, and none has ruled it out. The evidence settles a good deal about how we die and leaves open what, if anything, comes after.

What physically happens

The body shuts down in a fairly fixed order, and "the moment of death" is a legal marker placed on a gradual process. The timings below are typical figures from general medical knowledge, not from a single study, and vary with temperature, age and cause.

The last days and hours (expected deaths). Appetite and thirst fade. Sleep lengthens into unresponsiveness. Circulation pulls back to the core, so hands, feet and knees cool and mottle. Breathing turns irregular, with long pauses, and secretions the person can no longer clear produce the "death rattle". Hearing appears to be among the last senses to go: EEG work with unresponsive hospice patients has found the brain still reacting to sounds hours before death.

The minutes around cardiac arrest.

0 seconds: the heart stops pumping effectively. Blood pressure collapses.
About 10 to 20 seconds: consciousness is lost as the cortex runs out of oxygen. The surface EEG flattens within roughly 10 to 30 seconds.
About 1 to 5 minutes: brainstem reflexes fail. Breathing stops or becomes agonal gasping. Pupils fix and dilate.
Minutes in: neurons lose the energy to hold their electrical charge and discharge in a slow wave, called terminal spreading depolarization. This was long treated as the point of no return, though it is reversible if blood flow comes back soon enough.
Roughly 4 to 6 minutes without flow at normal body temperature: lasting brain injury becomes likely. Cold stretches this window a great deal.
Hours: individual cells die at different rates. Brain cells do not all die at once. In 2019 a Yale team restored cellular activity, though not organized brain activity, in pig brains four hours after death.

Two legal definitions. Circulatory death is the permanent stop of heartbeat and breathing. Brain death is the irreversible loss of all brain function, brainstem included, even while machines keep the heart beating.

After death.

Change	What it is	Typical timing
Pallor	Skin pales as circulation stops	Within 15 to 30 minutes
Cooling (algor mortis)	Body drifts toward room temperature	About 1 to 1.5 °F per hour at first
Lividity (livor mortis)	Blood settles to the lowest parts	Starts in 20 to 30 minutes, fixed by 6 to 12 hours
Stiffening (rigor mortis)	Muscles lock as their energy supply runs out	Starts at 2 to 6 hours, peaks near 12, fades by 24 to 48
Autolysis	Cells digest themselves with their own enzymes	Begins within minutes, visible over days
Putrefaction	Gut bacteria spread and break down tissue	Days to weeks

Some genes keep switching on for hours to days after death, a reminder that the organism and its cells do not die on the same clock.

What the dying brain has been recorded doing

The brain does not always go quietly, and three lines of measurement now show organized activity where older textbooks expected silence.

A surge after life support ends. A University of Michigan team reviewed EEG from four comatose patients who died after ventilators were removed. Two showed a burst of fast gamma waves within seconds to minutes, concentrated in a rear region tied to conscious perception and out-of-body sensations. The authors say plainly that this suggests conscious processing but does not show it, and that it could be a pathological by-product (Xu et al., PNAS 2023). The same lab found the surge in rats in 2013, and a 2022 case report caught a similar pattern in an 87-year-old man who died during a recording.
Brain waves during CPR. AWARE II monitored 567 in-hospital cardiac arrests at 25 US and UK hospitals. Despite severe oxygen shortage in the brain, near-normal EEG rhythms reappeared in some patients as late as 35 to 60 minutes into CPR (Parnia et al., Resuscitation 2023). CPR supplies some blood flow, so this is a resuscitated brain flickering, not a brain with no circulation at all.
Clarity before the end. Hospice research describes "terminal lucidity": people with advanced dementia or brain disease who become briefly clear and recognisable shortly before death. Separately, most dying hospice patients report vivid dreams or visions, often of dead relatives, and usually find them comforting. Neither has a settled explanation.

The limit on all of this is the same. The patients whose brains were recorded as they died could not be interviewed, and the survivors who were interviewed mostly had no EEG during the experience. Nobody has yet matched a specific reported experience to a specific recorded brain event.

Near-death experiences: how common, and what people report

Roughly 10 to 20 percent of people revived from cardiac arrest report a near-death experience (NDE), and the reports share a recognisable core across studies.

The benchmark study followed 344 consecutive cardiac arrest survivors in ten Dutch hospitals. Sixty-two (18%) reported some recollection from the period of clinical death, most of them shallow; 282 (82%) recalled nothing. Whether a patient had an NDE was not predicted by how long the arrest lasted, the drugs given, or prior fear of death (van Lommel et al., Lancet 2001). In AWARE II, of 28 survivors interviewed, 11 had some memory or perception from the arrest and 6 (21%) had a transcendent experience of the classic kind.

The recurring elements, in rough order of how often they appear:

Peace and absence of pain, often described as the strongest feeling of the person's life
A sense of leaving the body and watching from above
Movement through darkness or a tunnel toward a light
Meeting dead relatives, or a presence felt as loving and all-knowing
A life review, frequently felt from the point of view of the people one affected
A border or point of no return, and a decision or instruction to go back
Thought that feels faster and clearer than normal, and a memory that stays sharp for decades

Three further facts matter for interpretation.

You do not have to be dying. The same experiences occur in fainting, near-miss accidents, meditation and drug states, about as often as in true medical crises (Martial et al., Nature Reviews Neurology 2025).
Culture shapes the imagery. The feeling of peace and of meeting other beings shows up almost everywhere. The tunnel and the life review are much more common in Western accounts, and the figures met tend to match the person's own tradition.
The after-effects are large and lasting. Fear of death drops, interest in status and money falls, and concern for other people rises. In the Dutch study these changes were still growing eight years later, and were not seen in survivors without an NDE.

The good, the horrible, and the nothing

The three outcomes are very unequal in size: nothing is the usual result, peace is the usual experience, and terror is a real minority that was under-counted for decades.

Outcome	Share	What it is like	How people respond
Nothing recalled	About 80 to 90% of cardiac arrest survivors	A gap, like general anaesthesia: no time, no dark, no waiting	Usually no change in outlook
Pleasant experience	Most of those who report anything	Peace, light, reunion, life review, reluctance to return	Lasting loss of fear of death, changed values
Distressing experience	Roughly 14 to 23% of reported NDEs	Three forms, below	Often shame and silence, sometimes trauma

The horrible ones. Bruce Greyson and Nancy Evans Bush sorted distressing NDEs into three types (Bush and Greyson, Missouri Medicine 2014):

Inverse. The standard features occur, such as leaving the body or a life review, but are felt as terrifying or loss of control.
Void. An endless emptiness, total isolation, sometimes the conviction that nothing, including one's own life, was ever real. Many who have had it call this the worst type.
Hellish. Threatening beings, torment, falling, landscapes of suffering. This is the rarest of the three.

A Belgian analysis of 123 accounts found 17 distressing ones (14%), and found their memories just as detailed and vivid as blissful ones (Cassol et al., Memory 2019). Pooled figures from older studies run up to about 23%. The true rate is probably higher than reported, because people fear being judged and the standard scoring scale was built around pleasant feelings.

Two findings cut against the easy reading that good people get heaven and bad people get hell. Researchers have found no reliable link between the kind of life a person led and the kind of experience they had. And people respond in three ways: some treat it as a warning and change their lives, some explain it away, and some struggle for years to make sense of it.

The nothing. A blank can mean two different things, and current methods cannot tell them apart. Either no experience happened, or an experience happened and no memory was laid down. Memory formation is fragile and is knocked out by oxygen loss, sedatives and injury to the temporal lobes. Most dreams are forgotten the same way. So the majority's silence is not evidence that nothing was experienced, and it is not evidence that something was.

The case that the brain produces all of it

This is the case its defenders in neuroscience make, at full strength: every element of the NDE can be produced in a living brain by a known mechanism, so nothing beyond the brain is needed to explain it.

Each feature has a laboratory twin. Electrical stimulation near the junction of the temporal and parietal lobes produces out-of-body sensations. Fighter pilots blacking out in centrifuges report tunnel vision, floating and dreamlike scenes with loved ones. Ketamine and DMT produce experiences that score as NDEs on the standard scales. Endorphin release accounts for peace and painlessness.
The timing fits the surge. The gamma bursts and returning EEG rhythms show a brain capable of intense activity exactly when NDEs are said to occur. An experience that feels long can form in seconds, as dreams do, and may form while going under or coming back, not during the flat period.
The sleep link. People who have NDEs are more prone to REM intrusion, in which dream imagery and paralysis leak into waking. That predicts vivid, real-feeling scenes under stress.
A plausible evolutionary root. The 2025 NEPTUNE model ties these together: falling oxygen and rising carbon dioxide trigger a cascade of neurotransmitter release, and the result may descend from the feigned-death defence seen across animals (Martial et al.).
The content looks made by the person. Christians meet Christian figures and Hindus meet Hindu ones. Some people meet relatives who turn out to be alive. Experiences happen with no threat to life at all. These are what a brain drawing on its own memories would produce.
The critical test keeps failing. AWARE I and AWARE II placed hidden images where only a viewer near the ceiling could see them. Across both studies no patient has identified one. In AWARE II nobody named the visual target, and 1 of 28 recalled the audio cue.
Feeling real is not evidence. Psychedelic states and temporal lobe seizures also feel "more real than real". Certainty is a property of the experience, not a measurement of the world.

On this view the majority who recall nothing are the baseline, and the NDE is the last and most remarkable thing a brain does.

The case that consciousness survives

This is the case its defenders make, at full strength: the brain-based account explains fragments of the NDE in isolation and fails on the whole, and some reports should not be possible if the mind is only the brain.

Clear thought from a failing brain. Every other form of brain impairment, from low oxygen to concussion to drugs, produces confusion and patchy memory. NDEs are structured, lucid and remembered for life. Oxygen loss in particular does not fit: all cardiac arrest patients have it, yet only about one in five has an NDE, and the Dutch study found no medical variable that predicted who would.
Reported accurate perception. There are documented cases of patients describing details of their own resuscitation that they had no normal way to know. The best known are a Dutch patient who later identified the nurse who removed his dentures while he was in deep coma, and Pam Reynolds, who described instruments and conversation from a 1991 operation in which her body was cooled and her brain activity stopped. AWARE I recorded one patient whose account of events during about three minutes of arrest was checked against the record.
Knowledge the person lacked. Some experiencers report meeting someone they did not know had died. Researchers at the University of Virginia have collected such cases for decades.
Laboratory twins are loose copies. Stimulation-induced out-of-body sensations are brief and distorted. Drug states are usually recognised as drug states afterward. Neither reliably produces the life review, the encounter, or the permanent change in values.
The model has gaps its authors admit. Greyson and Pehlivanova argue that NEPTUNE left out evidence that contradicts it and does not address the defining features, and the model's own authors concede they cannot say which mechanisms are necessary or sufficient (UVA, January 2026).
The failed target tests prove little. So few patients survive, are interviewed and report leaving the body that the hidden images have barely been tested. Absence of a hit in a handful of cases is not a negative result.
Terminal lucidity. A mind returning in full through a brain destroyed by dementia is hard to square with the mind being that brain's output. It fits better with an old proposal from William James: the brain may transmit or filter consciousness, not generate it, the way a radio shapes a signal it does not create.
The unsolved base problem. Nobody can explain how brain tissue gives rise to inner experience at all. Until that is solved, the claim that it must end with the tissue is an assumption, not a finding.

On this view the NDE is a first glimpse of something that continues, and the people who recall nothing simply failed to carry the memory back.

What the traditions say happens spiritually

Human traditions give four broad answers, and the NDE evidence fits each of them partly and none of them cleanly.

Answer	Held by	What happens at death	Where NDE reports fit	Where they strain
Judgment and an eternal destination	Christianity, Islam, much of Judaism, Zoroastrianism	The soul leaves the body, is judged, and goes to paradise or punishment, in many versions after a waiting period and a resurrection	Light, a loving presence, a life review, hellish experiences	The review is usually felt as self-judgment with compassion, and experience type does not track belief or conduct
Rebirth	Hinduism, Buddhism, Jainism, Sikhism	Consciousness or karma carries into a new life. Tibetan Buddhism describes an in-between state, the bardo, with a clear light and then peaceful and wrathful visions made by one's own mind	Light first, then visions both blissful and terrifying; the idea that the mind shapes what it sees	Experiencers rarely report being directed to a new birth
Union or dissolution	Advaita Vedanta, mystical strands of most faiths, some modern spirituality	The separate self falls away into a larger awareness	Reports of oneness, boundless knowing, loss of the sense of self; the void	Most accounts keep a personal identity and meet personal relatives
Ending	Epicurean and modern secular thought	Experience stops. In Epicurus's argument, death is nothing to fear because no one is there to suffer it	The 80% or more who recall nothing; the anaesthesia-like blank	The vividness and after-effects of the minority

Three observations follow. The older texts and the modern reports resemble each other closely enough that either the texts record real NDEs from their own eras, or both come from the same human nervous system. The void experience, which devastates Western experiencers, is described in Buddhist and Hindu contemplative writing as a recognisable stage and not a punishment. And no tradition predicted the most stubborn finding, which is that who gets a blissful, a dreadful or a blank experience looks close to arbitrary.

What it all means

The evidence supports four conclusions with confidence, leaves one question open, and that open question is the one people care about most.

What is established.

Death is a process with a reversible early stretch. The line between alive and dead is set by medicine's ability to restart the body, and it has moved before and will move again.
The brain can be active, sometimes strikingly so, after the heart stops and during resuscitation.
NDEs are real experiences. They are not lies, and they are not ordinary confusion. They change people more durably than almost any other single event studied.
Dying, for the minority who remember it, is far more often peaceful than frightening. The body's last chemistry seems to work against suffering.

What is not established. Whether any experience occurs with no brain function at all. This is the whole question, and it rests on timing. An NDE is reported afterward by a brain that is working again. No study has yet pinned a verified perception to a moment when the EEG was flat and blood flow was absent. The strongest survival cases are single reports documented after the fact. The strongest brain-based evidence shows a mechanism that could produce the experience, which is not the same as showing that it did.

How the two cases stand against each other. The brain-based account is the more economical one. It needs nothing beyond known biology, it predicts the cultural shaping, the drug and fainting parallels and the failed target tests, and its gaps are of the kind that research usually closes. Its weak points are the lucidity of the experience, the unpredictability of who has one, and a small set of well-attested accurate reports. The survival account explains those weak points directly, but it has not yet produced a result under controlled conditions, and it has to explain why the content varies with culture. A fair summary is that the weight of evidence favours a brain origin for the experience, while the claim that consciousness therefore ends at death goes beyond what the evidence shows.

The distinction that often gets lost. A near-death experience is evidence about dying. It is not direct evidence about being dead. Everyone who has reported one came back, which means none of them reached the irreversible state. The NDE could be the brain's last act, or the doorway to something, and it would look the same from the inside either way.

What would settle it. One well-documented case of a patient correctly reporting a hidden target during a monitored flat EEG would overturn the brain-only view. Repeated failure to get one across hundreds of out-of-body reports, combined with EEG that tracks the experience moment by moment, would largely close the question the other way. A prospective study now under way follows patients from arrival in the resuscitation room, which is the first design able to line up brain recordings with later reports.

What it means for a person. Three things hold whichever case is right. The process of dying is likely to be gentler from the inside than it looks from the bedside. People near death may hear and understand more than they can show, so what is said to them matters. And the people who have been closest to the edge almost all come back with the same reordering of priorities: less concern for achievement and possessions, more for how they treated others. That finding does not depend on any view about an afterlife.

Sources

Figures from the studies below were taken from their abstracts and from institutional or press summaries found by search, not from the full papers. Timings in the physical section, the case reports, the cross-cultural and hospice findings, and the summaries of religious traditions are from general knowledge and should be read as approximate.