Near-Death Experiences: What the Research Actually Shows

People who nearly die, then survive, sometimes come back with a story. The stories are strange, and they repeat. That repetition is the first fact any honest account has to deal with.

Across published reports the same features show up again and again: a sense of being outside the body and looking down at it, moving through a dark space or tunnel toward a light, a feeling of peace instead of terror, meeting people who have died, and a rapid review of one's own life. Not everyone reports all of it. Most report a few pieces.

These are usually called near-death experiences, or NDEs. Researchers measure them with the Greyson NDE Scale, a 16-item questionnaire, and a score of 7 or higher counts as an NDE.^[6] The out-of-body sensation is one of the most common single features, present in about two-thirds (68 percent) of subjects in one review.^[7] How often NDEs happen depends on who you ask and how close to death they were. In a crowd-sourced sample of 1,034 adults across 35 countries, 10% reported an experience that met the threshold.^[6]

Two qualities come up in report after report and are harder to dismiss than the imagery. People describe the experience as more vivid and more coherent than ordinary waking life, not less, which is the opposite of a fading or confused brain. And they describe time as absent, with a whole life reviewed in what the clock recorded as seconds. Neither quality proves anything about an afterlife. Both are things a serious explanation has to account for rather than wave away.

Not every NDE is pleasant. A minority are frightening, marked by void, isolation, or a sense of judgment rather than peace. These distressing experiences are underreported, partly because people are less willing to describe them, and any honest picture has to include them alongside the light and the calm.

None of this is new. The tunnel and the light appear in medical writing going back more than a century, and in folk accounts long before that. What’s new is that researchers have started measuring NDEs before the person nearly dies instead of collecting them as anecdotes years later, and the measurements are more interesting than either side expected.

The strongest evidence comes from prospective studies, ones that set up the measurement in advance rather than gathering stories after the fact. Prospective design matters because it removes the biggest objection to the older anecdotes. When you enroll patients before the arrest, record the clinical facts as they happen, and interview survivors on a fixed schedule, you can’t be accused of cherry-picking dramatic stories or of a memory that grew in the retelling.

That’s why the prospective work carries more weight than a shelf of memoirs.

The landmark is Pim van Lommel's 2001 study in The Lancet. His team followed 344 patients across ten Dutch hospitals who were successfully resuscitated after cardiac arrest. 62 of them (18%) reported some memory from the period they were clinically dead, and 41 (12%) reported what the researchers called a core experience. The team went back at 2 years and again at 8 years to track what happened to those patients afterward.^[1]

Then came the AWARE studies, led by Sam Parnia. The first, published in 2014, screened 2,060 cardiac arrests across multiple hospitals. 140 survivors completed the first interview and 101 the follow-up. About 9% had experiences that fit the NDE pattern, and 2% described explicit, detailed recall of seeing and hearing their own resuscitation.^[2]

AWARE also tried something new. It placed images on shelves high in resuscitation rooms, visible only from the ceiling, to test whether patients who claimed to float above their bodies could actually report them.

AWARE-II followed in 2023. Across 567 in-hospital cardiac arrests, 53 patients survived and 28 were interviewed. 11 of the 28 (39%) recalled some perception, and about 21% described a transcendent experience of the kind usually called an NDE.^[3] No patient identified the hidden images, so the direct out-of-body test still hasn’t been passed.

Several physical mechanisms are on the table. None is proven, and they aren’t mutually exclusive.

• Oxygen starvation (anoxia). A brain running out of oxygen misfires. Low oxygen can produce tunnel vision, euphoria, and detachment. Pilots in high-G centrifuge training report similar sensations when blood drains from the brain.
• The temporal and parietal lobes. Electrically stimulating the junction of these regions can produce an out-of-body sensation on demand in the operating room, which ties the floating feeling to a specific piece of brain tissue.^[7]
• REM intrusion. The dreaming state of sleep can bleed into waking consciousness. Kevin Nelson's team found that people with NDEs report a lifetime history of sleep paralysis and sleep-related hallucinations far more often than matched controls.^[5] The crowd-sourced study found signs of REM intrusion in 47% of people with NDEs against 14% of those without.^[6]
• A surge at the end. In 2023, University of Michigan researchers recorded four comatose patients as life support was withdrawn. Two showed a sharp surge of gamma-band activity, the frequency tied to conscious processing, within seconds of the ventilator being removed, while the patients were dying.^[4] AWARE-II similarly caught spikes of normal-looking brain activity as far as 35 to 60 minutes into CPR, in brains with a mean oxygen saturation of 43%.^[3]

Here’s where honesty matters. Each mechanism explains part of the picture and leaves part unexplained.

The anoxia idea has a hole van Lommel pointed at directly. If oxygen starvation alone caused NDEs, nearly every resuscitated patient should have one, because they all lost brain oxygen. Only 18% did.^[1] Something about the person, not just the physiology of the arrest, seems to matter.

REM intrusion explains a predisposition, not the content. Temporal-lobe stimulation reproduces the out-of-body feeling but not the structured life review or the reported clarity. The gamma surge is real and striking, but it was recorded in two patients, and a burst of electrical activity isn’t the same thing as a verified conscious memory. The one claim that would settle the biggest question, seeing a target no dying brain could reach, hasn’t been confirmed in a controlled study despite two decades of trying.^[2][3]

There’s also a timing problem no one has solved. A memory reported after the heart restarts could have formed as the brain shut down, or as it came back online, or in some window in between. The subjective report says it happened during death. The clock can’t confirm that, and until it can, the strongest claims on both sides rest on faith rather than measurement.

Both camps overreach.

The believers point to lucid, structured experiences during periods when the brain looks too damaged to produce them, and conclude that consciousness outlives the body. That’s a leap the data doesn’t license. A vivid memory reported after resuscitation isn’t proof of what was happening during the flat stretch. The brain may build or timestamp the memory on the way back.

The debunkers point to anoxia, REM, and gamma surges and announce the case closed. That’s also a leap. No single mechanism accounts for the full phenomenon, the frequency, or the consistency of the content across cultures.

The defensible statement is narrow and unexciting, which is usually the sign it’s true. The experiences are real experiences. People aren’t lying, and they aren’t simply confused. What causes them, and what they mean, isn’t settled. Anyone who tells you the question is closed is selling either religion or its opposite.

One finding is solid enough to state plainly. People who have an NDE tend to stop being afraid of dying.

Van Lommel's team tracked this at 2 and 8 years. Patients who had an NDE showed a reduced fear of death that persisted, and it stayed lower than in resuscitated patients who had no such experience.^[1] Later work confirmed the pattern and began sorting out which features drive it, with encounters with mystical beings and a life review among the strongest predictors of reduced fear of death.^[8]

The aftereffect isn’t small or brief. In the follow-up studies the change in death anxiety was still measurable years later, and it often came with reordered priorities: less interest in status and money, more in relationships and time. People who have been to the edge and back describe living differently, and the studies bear that out even when the participants disagree about what happened to them.

This holds whatever the experience turns out to be. A hallucination that removes the fear of death changes how a person lives afterward as much as a glimpse of an afterlife would. If you want the plain version of what the end actually involves, read what dying actually looks like. If you want to understand why so many people can’t look at any of this, read why Americans are terrified of death.

No one can tell you what death itself is like; the reports we have come from people who didn’t stay dead. Two different experiences get blurred together here, and they aren’t the same thing.

The first is end-of-life visions, which happen to living people in their last days and weeks. Dying patients commonly describe seeing and talking to dead relatives and friends. In research on hospice patients by Christopher Kerr and colleagues at Hospice Buffalo, nearly 9 in 10 reported at least one such dream or vision, and patients more often found them comforting than distressing (Kerr et al., Journal of Palliative Medicine).

The second is the near-death experience, the tunnel-and-light reports that come from people resuscitated after cardiac arrest. That’s the subject of this article, and the evidence for it is covered above. The two get conflated constantly. Keep them apart: one is what a dying brain reports while still alive; the other is what a resuscitated brain reports after the heart stopped.

Some dying people do experience things that aren’t there, but "hallucinate" lumps together two very different situations. One is ordinary and usually calm; the other is a medical symptom that gets treated.

The calm kind is the end-of-life vision described above: seeing dead relatives, feeling reassured, and often lucid in between. Kerr's Hospice Buffalo research found that most hospice patients report these, and more found them comforting than distressing (Kerr et al., Journal of Palliative Medicine). The medical kind is terminal delirium: confusion, agitation, and restlessness driven by medications, organ failure, or infection. That’s a symptom, and the hospice team can treat it, so tell them.

Neither one is a near-death experience. NDEs are reported by people brought back after their heart stopped, not by people who are dying and still awake. This article covers the NDE research; end-of-life visions and delirium are a separate topic.