Converging evidence from brain-wave studies suggests that many people who appear unresponsive in their final hours still process sound. A landmark 2020 electroencephalography study found that most dying patients who could no longer react to their surroundings nonetheless showed auditory brain responses similar to those of young, healthy volunteers, sometimes just hours before death.1PubMed Central. Electrophysiological evidence of preserved hearing at the end of life The question of whether a dying person can hear you is more nuanced than a flat yes or no, but the science leans strongly toward yes, and the implications for how we behave at the bedside are real.
What the EEG Evidence Actually Shows
The most direct evidence comes from a study conducted at the University of British Columbia, in which researchers placed EEG caps on both hospice patients in their final hours and a control group of healthy young adults. Both groups listened to sequences of tones designed to test whether the brain could detect changes in pitch and pattern. The researchers were looking for specific brain-wave signatures: one called the mismatch negativity response, which fires automatically when the brain notices a deviation in a sound pattern, and higher-order responses that suggest deeper cognitive processing of what was heard.
Most of the unresponsive patients produced the automatic detection response, indicating their auditory cortex was still registering changes in sound. Some also showed the higher-order responses, which in healthy people are associated with consciously noticing that something has changed. The researchers concluded that the auditory systems of these patients were responding in ways strikingly similar to the healthy controls, even though the patients showed no outward signs of awareness.1PubMed Central. Electrophysiological evidence of preserved hearing at the end of life The study could not prove that the patients consciously experienced what they heard, but it demonstrated that the machinery for hearing was still running.
Why Hearing Seems to Be the Last Sense to Go
The auditory system has a structural advantage over other senses when the body is shutting down. Sound enters the ear, gets converted into electrical signals, and travels along a pathway that runs through the brainstem before it reaches the cortex. The brainstem is the oldest, most deeply embedded part of the brain, and it is among the most resistant to the effects of low oxygen and failing circulation. Vision, by contrast, depends heavily on the occipital cortex, a region more vulnerable to metabolic decline. Touch and proprioception degrade as peripheral circulation collapses. But the brainstem auditory pathway can keep functioning even when higher brain regions are struggling.
Research on patients with the lowest possible score on the Glasgow Coma Scale, a measure of consciousness, offers indirect support for this. Even in some of these deeply unresponsive patients, auditory brainstem responses could still be recorded. When those brainstem responses were absent or abnormal, the patients were significantly more likely to die, suggesting that the auditory brainstem pathway is one of the last neural systems to fail.2PubMed. Auditory brainstem response: prognostic value in patients with a score of 3 on the Glasgow Coma Scale In other words, as long as the brainstem is alive, there is plausible biological infrastructure for hearing to persist.
The UBC researchers put it plainly: hearing may indeed be one of the last senses to lose function as humans die.1PubMed Central. Electrophysiological evidence of preserved hearing at the end of life
What Anesthesia Research Tells Us About Unconscious Hearing
The dying brain is not the only context in which hearing outlasts apparent consciousness. General anesthesia provides a useful comparison, because anesthetized patients are, by design, rendered unconscious and unresponsive. Yet a systematic review and meta-analysis of studies on auditory perception during surgery found that patients can perceive and process sounds during clinically adequate general anesthesia, even without forming any explicit memories of the experience.3PubMed Central. Perception of auditory stimuli during general anesthesia and its effects on patient outcomes: a systematic review and meta-analysis
The evidence for this comes from implicit memory tasks. After waking up, patients who had heard specific words or sounds during surgery were more likely to complete word stems with those words, even though they had no conscious memory of hearing anything. Roughly four in ten of the studies that tested for this kind of unconscious memory found positive results.3PubMed Central. Perception of auditory stimuli during general anesthesia and its effects on patient outcomes: a systematic review and meta-analysis Even more striking, patients who heard music played during their surgery reported less postoperative pain and needed fewer opioids afterward, while positive verbal suggestions played during surgery did not have the same effect.
The anesthesia literature matters here because it establishes a principle: even when a person cannot respond and will not remember, the brain can still take in and be influenced by what it hears. The mechanism is different from natural dying, since anesthetic drugs suppress consciousness chemically rather than through oxygen deprivation, but the takeaway is similar. Unresponsiveness is not the same as deafness.
Surges of Brain Activity Near Death
One of the more surprising findings in recent years is that the dying brain does not simply wind down like a battery running flat. In both animal studies and a small number of clinical observations, researchers have detected brief surges of high-frequency electrical activity in the brain around the time of death. This activity falls within the gamma frequency band, a range associated in healthy brains with conscious awareness, attention, and the binding together of sensory information.4PubMed Central. The gamma-band activity model of the near-death experience: a critique and a reinterpretation
These surges are paradoxical. They occur at a time when cortical electrical activity is otherwise at a very low ebb, meaning the overall brain is going quiet while this specific band of activity flares. Some researchers have proposed that these bursts could be the neural basis for near-death experiences, though this idea remains debated. The gamma-surge model is suggestive rather than proven: it offers a potential mechanism for why some people who survive cardiac arrest report vivid perceptions during the period when their hearts had stopped, even though only a few survivors of cardiac arrest recall anything from the event at all.5Medical Research Archives. Consciousness during Cardiopulmonary Resuscitation: A Case Report
Whether or not these gamma surges produce something like conscious experience, they reinforce a broader point: the transition from life to death is not a smooth, linear decline in brain activity. There are pockets of intense neural processing that persist or even spike during the process. A brain capable of producing gamma-band surges in its final moments is, at the very least, more active than most people assume.
Terminal Lucidity and What It Suggests About Awareness
Some of the most compelling anecdotal evidence that dying people remain aware comes from a phenomenon called terminal lucidity, or paradoxical lucidity. This refers to unexpected episodes where patients with severe cognitive impairment, including advanced dementia, suddenly become alert, oriented, and communicative shortly before death. The most dramatic historical case may be that of Anna Katharina Ehmer, a 26-year-old woman with severe mental disabilities who had reportedly never spoken a single word during her entire life, yet was described as singing dying songs for half an hour before she died.6PubMed. The death of Anna Katharina Ehmer: a case study in terminal lucidity
Modern research has moved beyond case reports. A multi-site prospective study of patients with moderate to severe dementia found that about 62% of the 151 participants exhibited at least one episode consistent with lucidity. These episodes included appropriate orientation to events, return of old memories, return of functional abilities, and meaningful nonverbal communication.7Innovation in Aging. A Multi-Site Prospective Study of Paradoxical Lucidity in Moderate to Severe Dementia A smaller subset, about 4%, showed what the researchers classified specifically as terminal lucidity, meaning the episode occurred near death and included features consistent with near-death experiences.
Terminal lucidity does not directly prove that all dying people can hear you. But it demonstrates that the dying brain is capable of far more complex processing than its outward appearance suggests. If a person with severe dementia who has not spoken coherently in years can suddenly recognize family members and hold a conversation, it becomes much harder to assume that an unresponsive dying person is necessarily unaware of voices in the room. Healthcare professionals who witness these episodes often describe them as profoundly surprising, and pilot studies have begun systematically cataloguing their features, including the degree of both expressive and receptive communication involved.8PubMed Central. Reports About Paradoxical Lucidity from Health Care Professionals: A Pilot Study
Music at the Bedside and What It Changes Physiologically
If dying patients can still hear, then what they hear matters. This is where music therapy research provides some of the most practical evidence. A meta-analysis covering studies of terminally ill patients found that music therapy significantly reduced pain and improved quality of life compared to standard palliative care. It also reduced anxiety and depression, though it did not significantly improve physical status or fatigue.9PubMed. The Effectiveness of Music Therapy for Terminally Ill Patients: A Meta-Analysis and Systematic Review
The physiological effects are measurable even in patients who cannot tell you what they are feeling. In one hospice study, heart rates dropped from about 86 beats per minute before music therapy to about 77 beats per minute afterward, and respiratory rates fell from roughly 20 breaths per minute to about 15. Skin temperature in the extremities rose slightly, which the researchers interpreted as a sign of relaxation through improved blood flow to the hands and feet.10PubMed Central. Music Therapy in Hospice and Palliative Care: a Review of the Empirical Data
A separate study examining the effects of music therapy on vital signs and anxiety in terminally ill patients confirmed the heart rate finding and added that blood oxygen saturation improved significantly after sessions. Blood pressure dropped meaningfully in patients whose readings had been elevated beforehand, while patients with already-normal blood pressure did not show a significant change. Anxiety scores fell sharply.11Journal of Hospice & Palliative Medical Care. Effects of Music Therapy on Vital Signs and Anxiety: A Study with Terminally Ill Patients These findings are consistent with the idea that the auditory system is not only intact but functionally connected to the autonomic nervous system: what the dying person hears can change their heart rate, breathing, and stress level, even when they cannot say a word.
What to Say and What Not to Worry About
For anyone sitting at the bedside of a dying loved one, the practical question is less about whether they can hear you and more about what you should do with that possibility. Hospice workers have long encouraged families to talk to unresponsive patients, to say what they need to say, and to avoid conversations in the room that they would not want the patient to hear. The EEG and music therapy evidence supports this practice as more than wishful thinking.
A common concern is whether saying goodbye or expressing grief will cause distress. There is no evidence for this. The physiological data from music therapy studies point in the opposite direction: calming, familiar auditory input appears to reduce stress markers in dying patients. Hearing a loved one’s voice is, by almost any reasonable interpretation, a form of calming, familiar input. Families who worry about “bothering” a dying person by talking to them can be reassured that the far more likely effect is comfort, not distress.
Another question family members often have is whether they need to be physically present and speaking continuously. The answer is no. The studies that measured auditory processing used relatively brief stimulus presentations. Nobody is suggesting that the dying brain needs constant narration. Silence, gentle music, the ordinary sounds of family life in the room: these are all auditory environments that the evidence suggests can be perceived, and none of them are harmful. The key insight is simply that the room is not as private as it might feel when the patient appears unconscious. Difficult family conversations, medical discussions about prognosis, or emotional arguments are worth having in the hallway rather than at the bedside.
The Physiological Landscape of the Final Days
Understanding what else is happening in the body during the final days provides context for why hearing seems disproportionately preserved. A study of advanced cancer patients tracked vital signs during the last three days of life. Blood pressure and oxygen saturation dropped significantly during this window, while temperature crept up slightly. Heart rate and respiratory rate, somewhat surprisingly, did not change significantly over the same period.12PubMed Central. Variations in Vital Signs in the Last Days of Life in Patients With Advanced Cancer
The drop in oxygen saturation is relevant to hearing because the brain is sensitive to hypoxia. But as the brainstem auditory pathway data suggest, the structures needed for basic auditory processing are among the most hypoxia-resistant in the nervous system. Falling blood pressure reduces perfusion to the whole body, but the brainstem, being essential for sustaining life, tends to receive preferential blood flow until late in the dying process. This is part of why auditory responses can persist even as other systems fail.
The elevated heart rate that was associated with impending death in the vital signs study also has implications. A faster heart rate in the final days may reflect sympathetic nervous system activation, the body’s stress response. If the autonomic nervous system is still active enough to accelerate the heart, it is plausible that auditory input could modulate it. The music therapy findings, where heart rate consistently drops during sessions, suggest exactly this kind of interaction.
How End-of-Life Presence Affects the People Left Behind
The question of whether a dying person can hear is also, in a very real sense, a question about the people doing the talking. Research on bereaved family caregivers has found a significant link between how caregivers perceive the quality of end-of-life care and their subsequent grief. Caregivers who felt that the dying person received better care, including better psychological support and better symptom control, reported lower levels of grief afterward. Similarly, caregivers who perceived a higher quality of dying and death, including better preparation for death and a sense of transcendence, also grieved less intensely.13PubMed Central. Quality of end-of-life care, quality of dying and death, and grief in bereaved family caregivers
This research does not directly measure the effect of talking to a dying person on the speaker’s grief outcomes. But it establishes that how the death unfolds matters to the survivors. Families who feel they were present, who said what they needed to say, and who treated the dying person as still capable of receiving their words are, in essence, shaping the quality of their own experience of the death. If the dying person can also hear and derive comfort from those words, the benefit runs in both directions.
The evidence is strong enough that the old hospice advice, “talk to them as if they can hear you,” deserves an upgrade. It is not just a kindness or a hedge. It is consistent with what the neuroscience, the physiology, and the clinical data all point toward: the auditory system keeps working far deeper into the dying process than most people expect, and what reaches the ear may well reach the person behind it.
The Limits of What We Know
For all the converging evidence, genuine uncertainty remains. EEG can tell us that the auditory cortex is responding to sound, but it cannot tell us whether the person is experiencing anything subjective. A brain that generates a mismatch negativity response is detecting a change in the auditory environment, but that detection can, in theory, happen without consciousness. The higher-order responses seen in some patients are more suggestive of conscious processing, but even they are not proof.
The gamma-band surges near death are intriguing but have been observed in very few clinical cases. The animal data is more robust, but extrapolating from rodent brains to human experience is always uncertain. Terminal lucidity, while now documented prospectively, still lacks a clear mechanistic explanation. Nobody knows why a brain ravaged by years of neurodegeneration can suddenly produce coherent speech and accurate memories. Until that mechanism is understood, terminal lucidity remains more of a phenomenon to be catalogued than a tool for making predictions about any individual patient.
There is also variability among patients. Not every unresponsive person in the UBC study showed auditory responses. Some showed only the basic automatic detection, some showed higher-order responses, and presumably some showed nothing at all, though the published framing emphasizes the positive findings. The degree of sedation, the underlying cause of unresponsiveness, and the individual patient’s neurological reserve all likely affect whether hearing is preserved in any given case. For families at the bedside, the honest framing is: we cannot guarantee your loved one hears you, but the balance of evidence suggests it is more likely than not, and there is no downside to assuming they can.