The earliest reliably detected brain waves in a human fetus appear around 24 to 28 weeks of gestation, coinciding with the formation of the first functional connections between deep brain structures and the cerebral cortex. Before that window, individual neurons fire spontaneously, but the coordinated electrical patterns recognizable as “brain waves” on a recording depend on circuitry that simply has not been built yet. The story of when those waves appear, what they look like, and what they mean is more layered than a single gestational-age number can convey.
What Has to Happen Before Brain Waves Can Exist
Brain waves are not produced by individual nerve cells acting alone. They emerge when large populations of neurons fire in coordinated rhythms, and that requires physical wiring between regions of the brain. The critical event in the fetal brain is the arrival of nerve fibers from the thalamus, a relay hub deep in the brain, into the outer layer of the cortex where higher processing occurs. These thalamocortical fibers do not simply grow straight to their destination. They pause for weeks in a temporary structure called the subplate zone, a kind of staging area just below the cortical surface. Once the cortex is ready, the fibers push through and form the first working synapses in the cortical plate, creating the hardware necessary for coordinated electrical activity.1PubMed. The development of the subplate and thalamocortical connections in the human foetal brain
This process unfolds gradually. In the visual cortex, for example, functional thalamocortical connectivity increases significantly from about 19 to 40 weeks, but the steepest rise occurs around 29 to 31 weeks, when thalamocortical fibers are actively invading the cortical plate.2Cerebral Cortex. Fetal development of functional thalamocortical and cortico–cortical connectivity In the somatosensory cortex, the region that processes touch, a similar penetration of fibers into the cortical plate creates the first distinct band of activity there.3PubMed Central. Growth of Thalamocortical Fibers to the Somatosensory Cortex in the Human Fetal Brain Different cortical regions mature on slightly different schedules, so the fetal brain does not “switch on” all at once. Rather, it comes online region by region over several weeks in the late second and early third trimester.
Spontaneous Activity Before True Brain Waves
There is a distinction worth drawing between the coordinated oscillations we call brain waves and the much earlier, more primitive electrical activity produced by individual neurons. Electrophysiology experiments on human fetal cortical tissue have shown that subplate neurons express functional receptors for the brain’s main excitatory and inhibitory signaling chemicals as early as 20 weeks of gestation. These neurons produce spontaneous bursts of firing that resemble, in miniature, the “up and down states” seen in the adult cortex during sleep.4PubMed Central. Spontaneous electrical activity in the human fetal cortex in vitro But actual synaptic connections between neurons at this stage are rare. The activity is internally generated and largely disconnected from one another. It is a sign that the raw components are coming alive, not that a working network is in place.
This matters because claims about “brain activity” at very early gestational ages sometimes conflate single-neuron firing with organized brain waves. The former appears well before viability. The latter requires the thalamocortical wiring described above and does not become detectable until the late second trimester at the earliest.
The 28-Week Milestone
The clearest answer to “when can we actually record fetal brain waves?” comes from studies using non-invasive magnetoencephalography, or fMEG. This technique picks up the tiny magnetic fields that electrical currents in the brain produce, measured through sensors placed on the mother’s abdomen. Using fMEG, researchers have identified specific brain activity patterns in fetuses starting at 28 weeks of gestation. In a study of 30 fetuses, 60 percent showed at least one recording with discontinuous activity, about a quarter showed sharp transients, and roughly the same fraction showed a pattern called delta brush activity.5PubMed. Non-invasive detection and identification of brain activity patterns in the developing fetus
fMEG remains the only non-invasive method capable of directly recording evoked brain responses and spontaneous brain activity from a fetus still in the womb. Both auditory and visual evoked responses have been successfully captured this way.6PubMed Central. Fetal Magnetoencephalography – Achievements and Challenges in the Study of Prenatal and Early Postnatal Brain Responses: A Review The technology is expensive and not widely available in clinical settings, but it has been instrumental in pinning down the timeline of when organized fetal brain activity begins.
Delta Brushes and What They Tell Us
If you could look at an EEG tracing from a premature infant born at 28 or 30 weeks, the most recognizable feature would be the delta brush: a slow wave with a burst of faster oscillations riding on top of it, a bit like a wave cresting with small ripples. Delta brushes are the hallmark electrical pattern of the premature brain and serve as a key marker of neural maturation.7PubMed Central. Characteristics and clinical significance of delta brushes in the EEG of premature infants They are thought to reflect the activity of newly forming thalamocortical synapses as the subplate zone gradually hands off its duties to permanent cortical circuits.
The fast oscillatory component of a delta brush typically runs in the range of about 15 to 20 Hz. Interestingly, premature infants show more powerful delta brush activity than full-term newborns, consistent with the idea that delta brushes are a feature of an immature brain actively building connections rather than a permanent part of neural function.8PubMed. Rapid oscillatory activity in delta brushes of premature and term neonatal EEG As the brain matures toward term, delta brushes become less prominent and eventually disappear, replaced by more adult-like rhythms. Their presence in expected amounts at expected gestational ages is a reassuring sign; their absence or abnormality can raise flags about brain development.
Fetal Sleep States Emerge With Brain Waves
One of the most tangible consequences of developing brain wave activity is the emergence of sleep states. The cycling between active sleep (analogous to REM) and quiet sleep (analogous to non-REM) depends on coordinated electrical patterns that only become possible once thalamocortical networks are functioning. In human fetuses, distinct REM and non-REM periods become clearly distinguishable between about 28 and 31 weeks of gestation.9Early Human Development. A study on the development of sleep-wakefulness cycle in the human fetus
Full maturation of these sleep states, however, takes considerably longer. A large cross-species review found that in humans, organized cycling between active and quiet sleep does not fully mature until around 95 percent of gestation, which means very close to term.10Sleep. Fetal sleep: a cross-species review of physiology, measurement, and classification Fetal eye movements track this maturation: eye movement density shows critical transitions at 28 to 29 weeks and again at 36 to 37 weeks, roughly mirroring the two-phase development of sleep states.11PubMed Central. Eye movement activity in normal human fetuses between 24 and 39 weeks of gestation
This means a fetus at 28 weeks has rudimentary sleep cycling, but the crisp alternation between active and quiet states that a full-term newborn shows is still weeks away. The brain waves that define each state are present but immature, gradually sharpening as more cortical connections come online.
When the Fetal Brain Responds to Sounds and Light
Spontaneous brain waves tell you the network exists. Evoked responses tell you it is processing information from the outside world. Using functional MRI, researchers have detected cortical activation in response to sound in fetuses as young as 33 weeks. The finding confirmed that sound processing at this stage goes beyond mere reflexive subcortical responses and involves the cortex itself.12NeuroImage. Fetal cortical activation to sound at 33 weeks of gestation: A functional MRI study
Auditory evoked responses measured by fMEG have been recorded even a few weeks earlier, though the signals are small and variable. The general pattern is that sensory evoked activity becomes more reliable and faster as the third trimester progresses, reflecting the ongoing maturation of thalamocortical pathways. Visual evoked responses have also been recorded in utero, though the fetal visual system matures later than the auditory one, and in-utero light stimulation is harder to standardize than sound.
An intriguing extension of this work has shown that maternal metabolism directly influences how quickly the fetal brain responds to stimuli. After mothers drank a glucose solution, fetal auditory response times sped up from roughly 297 milliseconds at baseline to about 235 milliseconds within an hour. Even more striking, fetuses of insulin-resistant mothers responded more slowly to sounds than fetuses of insulin-sensitive mothers, suggesting that metabolic conditions during pregnancy can shape fetal brain function in real time.13PubMed. Maternal insulin sensitivity is associated with oral glucose-induced changes in fetal brain activity
Clinical Uses of Fetal Brain Wave Monitoring
Attempts to use fetal brain recordings clinically have a surprisingly long history. Studies on the feasibility of monitoring fetal EEG during labor began in the early 1940s. By the 1970s, researchers were reporting clear diagnostic and prognostic benefits from intrapartum fetal EEG monitoring, yet the technology never moved into routine use and has remained, in the words of one systematic review, “a curiosity.”14PubMed Central. Monitoring Fetal Electroencephalogram Intrapartum: A Systematic Literature Review The main barriers are practical: the fetal skull is surrounded by amniotic fluid, maternal tissue, and the uterine wall, all of which distort or attenuate electrical signals. Magnetic signals (used in fMEG) pass through tissue more cleanly, but the equipment is bulky, expensive, and requires a magnetically shielded room.
Where fetal brain monitoring has proven valuable is in detecting problems. Fetal seizures, though rare, can sometimes be identified before birth. In one reported case, repetitive rhythmic jerking movements at 32 weeks were confirmed as seizures by fMEG, which revealed immature patterns consistent with encephalopathy, including periods of severe bilateral suppression alternating with low-amplitude activity and bursts lasting up to six seconds that coincided with body movements.15Journal of Clinical Neurophysiology. Magnetoencephalography as a Novel Tool for Prenatal Diagnosis of Fetal Seizures In another case, tonic-clonic movements of the fetal trunk and extremities were seen on ultrasound at 30 weeks after the mother reported rapid, repetitive fetal movement beginning around 28 weeks.16PubMed Central. Prenatal diagnosis of fetal seizure: a case report
Growth restriction caused by placental insufficiency is another area of concern. When the placenta cannot deliver enough oxygen, the fetus redirects blood flow to protect the brain, a response called brain sparing. Even with this compensatory mechanism, chronic oxygen deprivation can still affect brain structure and development.17PubMed Central. The consequences of fetal growth restriction on brain structure and neurodevelopmental outcome Monitoring fetal brain activity could, in theory, help clinicians detect when the brain is under stress before visible structural damage occurs. The same systematic review noted that the detection of sharp waves on fetal EEG during labor correctly classified 76 percent of patients in terms of neurological outcome at one year.14PubMed Central. Monitoring Fetal Electroencephalogram Intrapartum: A Systematic Literature Review
What Animal Studies Add to the Picture
Much of what we know about fetal brain electrical development comes from animal models, particularly fetal sheep, which allow continuous in-utero recording that is impossible in human pregnancies. The fetal sheep brain develops organized sleep states from about 115 to 120 days of gestational age (term in sheep is around 150 days), corresponding to roughly 77 to 80 percent of gestation.18PubMed. Developmental changes in the complexity of the electrocortical activity in foetal sheep This is earlier in relative terms than the human timeline, where full sleep state organization does not mature until about 95 percent of gestation.
Sheep studies have also shown how chronic oxygen deprivation alters the maturation of brain electrical activity, lending support to concerns about growth-restricted human fetuses.19PubMed. Maturational changes and effects of chronic hypoxemia on electrocortical activity in the ovine fetus Classic comparative work across species reveals a striking principle: animals born at an advanced stage of development, like guinea pigs and chickens, show spontaneous cortical electrical activity before birth, while animals born helpless, like rats and pigeons, do not develop detectable cortical activity until days after birth.20The Japanese Journal of Physiology. COMPARATIVE STUDIES ON THE DEVELOPMENT OF EEG Humans fall somewhere in between: brain waves emerge before birth, but they are far from mature at term.
An interesting side note from animal work: administering a barbiturate sedative to a pregnant guinea pig changed the fetal electrical patterns in close parallel with changes in the mother’s brain.21Experimental Neurology. Fetal and maternal electroencephalography in the guinea pig This underscores how permeable the placental barrier is to substances that affect neural activity, a finding with obvious implications for anesthesia and medication use during pregnancy.
Brain Waves and the Question of Consciousness
The detection of fetal brain waves inevitably raises the question of whether and when a fetus might be conscious. This is where neuroscience meets philosophy, and the evidence gets genuinely complicated. One influential analysis placed “brain birth,” by analogy with brain death, at 24 to 28 weeks of gestation, based on the EEG characteristics of premature infants and the developmental milestones required for coordinated cortical function. But the same paper cautioned that the gradualness of brain development makes the analogy with brain death problematic; brain death is a sudden loss of function, while brain maturation is a slow accumulation of it.22PubMed Central. Brain birth and personal identity
A separate philosophical analysis that looked more closely at the specific EEG patterns associated with conscious experience in adults arrived at a later estimate, concluding tentatively that a fetus becomes conscious at about 30 to 35 weeks after conception.23PubMed. When did you first begin to feel it? — locating the beginning of human consciousness The reasoning hinged on the maturation of the chemical and electrical pathways that underpin the kind of cortical activity associated with sensory experience in adults. At 28 weeks the hardware exists, but the patterns of activity that correlate with awareness in older humans are not yet reliably present.
The honest summary is that brain waves at 24 to 28 weeks reflect functioning neural circuits, but the leap from “the cortex is electrically active” to “the fetus is experiencing something” is one that current science cannot make with confidence. The subplate-driven activity of the late second trimester may be more akin to the brain rehearsing its circuitry than to a mind perceiving the world. The thalamocortical connections are in place, the patterns are detectable, and they grow more complex week by week, but exactly when that complexity crosses into subjective experience remains an open question. The connection between thalamocortical wiring and the generation of transient electrical phenomena may represent an early, provisional network in the developmental history of consciousness, but not consciousness itself.1PubMed. The development of the subplate and thalamocortical connections in the human foetal brain
Why the Range Matters More Than a Single Number
People searching for a definitive week number will find it frustrating that the answer spans roughly 24 to 35 weeks depending on what you mean by “brain waves.” If you mean any coordinated electrical activity at all, the subplate neurons are firing spontaneously at 20 weeks in tissue preparations, though this is not detectable in a living fetus. If you mean patterned activity recognizable on a recording, 28 weeks is the most commonly cited starting point. If you mean the mature brain wave patterns associated with distinct sleep states and reliable sensory processing, you are looking at 35 weeks or later.
The reason for this spread is not measurement imprecision. It reflects the biological reality that brain development is not a switch that flips. Different cortical regions wire up on different schedules. The thalamocortical connectivity that drives brain waves increases along a sigmoid curve with inflection points scattered across several weeks depending on the brain region involved.2Cerebral Cortex. Fetal development of functional thalamocortical and cortico–cortical connectivity The subplate zone gradually dissolves as permanent associative pathways take over, activating large-scale neural networks that support more complex electrical patterns.24Journal of Neuropathology & Experimental Neurology. Fundamentals of the Development of Connectivity in the Human Fetal Brain in Late Gestation Each of these transitions adds another layer to the brain’s electrical repertoire, making “brain waves detected” less a single event and more a months-long crescendo that starts faintly around the end of the second trimester and does not reach its full prenatal form until close to birth.