Is White Noise Bad for You? Risks and Safe Use

Two numbers, not a colour

Is white noise bad for you? Two numbers answer that, and neither of them mentions white noise. The US National Institute on Deafness and Other Communication Disorders states that sounds at or below 70 dBA are unlikely to cause hearing loss even after long exposure, and that long or repeated exposure at or above 85 dBA can cause it. Those limits apply to any sound. The spectral shape does not enter into it, so the question is really one about level and duration — with one genuine exception, infants, covered below.

Hearing damage accumulates from sound energy over time, which is why both numbers in the threshold come paired with a sense of duration. Below 70 dBA, long exposure is not the concern. Above 85 dBA, the time needed to do damage shortens as the level rises. For context, the same source puts normal conversation at 60 to 70 dBA.

What the damage consists of is also specific. The NIDCD describes noise-induced hearing loss as the damage and eventual death of the hair cells in the cochlea, and notes that in humans those cells do not grow back. That is the reason the guidance is framed as prevention rather than treatment: there is no later intervention that restores the cells, so the only control available is the level and the length of the exposure.

For overnight or all-day background sound, the practical guidance follows directly: keep it low enough that it does not compete with conversation. A level you have to raise your voice over is a level worth turning down, and nothing about continuous background sound requires it to be loud to work — masking depends on raising the floor relative to the intrusions, not on volume for its own sake.

One caveat concerns the delivery rather than the sound. In-ear earbuds place the driver in the ear canal, which makes the level at the eardrum harder to judge than it is with a speaker across the room, and wearing them for eight hours a night can irritate the canal. A speaker at a distance is the easier thing to get right.

Dose, not just loudness

The thresholds above are the short version. The standards that govern listening through headphones work in terms of a dose, which combines level with hours, and that framing is the more useful one for anyone who leaves a sound running all night.

The WHO-ITU standard for safe listening devices, published in 2019 and built on the ITU-T H.870 guidelines, sets a weekly allowance rather than a volume limit: 80 dB for 40 hours a week for adults, and a more conservative 75 dB for 40 hours a week for children. The allowance works on an equal-energy rule, so the permitted time halves for every 3 dB of extra level — 20 hours at 83 dB, 10 hours at 86 dB, and so on down. The standard asks devices to track that dose and warn the user as the allowance is used up.

Two things follow for overnight use. First, a dose model is the right way to think about a sound that runs for eight hours: level and duration trade against each other, so a quiet sound for a long time can carry the same energy as a louder one for a short time. Second, the rule runs in both directions, and running it downwards is reassuring. The 40 hours are 40 hours at 80 dB; every 3 dB below that doubles the time, so a quiet background sound buys a far larger allowance: about 400 hours a week at 70 dB, and about 1,280 hours at 65 dB. Eight hours a night across seven nights is 56 hours, which at the levels this page recommends is a small fraction of the week’s allowance rather than a brush with it.

So the reason to prefer a speaker is not that the allowance is being approached. It is that nobody has measured the hearing of people who sleep in headphones, and that an in-ear driver makes the level at the eardrum the one quantity in this whole calculation you cannot judge by ear. A speaker a few feet away keeps the level both lower and estimable; earbuds keep it neither.

Infants: the one case with measurements behind it

White noise is widely used to settle babies, and this is the area where specific evidence exists and where it is not reassuring. A 2014 study in Pediatrics measured the output of fourteen infant sleep machines at three distances and found mean maximum levels of 79 dBA at 30 cm, 71 dBA at 100 cm, and 63 dBA at 200 cm. Three machines exceeded 85 dBA at the closest distance. All fourteen exceeded 50 dBA — the recommended noise limit for hospital nurseries — when measured at 30 cm.

Vertical decibel scale from 30 to 110 A-weighted decibels, divided into three bands: below 70 dBA, unlikely to harm hearing; 70 to 85 dBA, caution; above 85 dBA, harm with long exposure. Marked thresholds at 85, 70 and 50 dBA, the last labelled as the nursery limit. Three points to the right show mean maximum infant sleep machine output: 79 dBA at 30 centimetres, in the caution band; 71 dBA at 100 centimetres, just inside caution; and 63 dBA at 200 centimetres, inside the safe band. A-weighted decibels 85 70 50 110 30 Harm with long exposure Caution Unlikely to harm hearing Below 50: nursery limit 79 dBA at 30 cm 71 dBA at 100 cm 63 dBA at 200 cm Infant sleep machines, mean maximum output
The two thresholds that answer the question, and three measurements against them. Hugh et al. measured the mean maximum output of 14 infant sleep machines at 30, 100 and 200 cm; distance alone moves the same device from the caution band into the safe one. Thresholds from NIDCD; the 50 dBA line is the hospital nursery limit cited in that study.

The figure makes the practical point better than the prose can: the same machine at the same setting moves from the caution band to the safe band on distance alone. Placing the device across the room rather than beside the crib, and running it quietly rather than at maximum, is what brings the exposure down. Infants cannot report that something is too loud, and their exposure is measured in whole nights, which is why the margin here should be larger than an adult would bother with. For any question about a specific child, a paediatrician is the right source.

The study’s own recommendations to families are three: place the machine as far from the infant as possible and never in the crib or on a crib rail, play it at a low volume, and run it for a limited time rather than all night. The authors also asked manufacturers to cap maximum output and fit a timer. The first two recommendations follow directly from their measurements. The third does not follow from them. The study measured output level at three distances; it did not measure anything over time, and the authors’ concern about duration rests on a separate argument about auditory development. One of the authors told NPR that no large study of children raised with these machines has been done, which is the state of the evidence on duration as well. It is also the one recommendation clinicians argue about in public. In CNN’s coverage of the study, the paediatrician Harvey Karp called the short-duration advice “not supported by the data in the study” and recommended running the sound for the whole of sleep, on the grounds that stopping it causes more waking in the middle of the night; Blake Papsin, the study’s senior author and an otolaryngologist whose clinical work is the surgical rehabilitation of hearing loss, maintained the opposite. Both have something at stake worth naming: Papsin’s field is the hearing the study is concerned about, and Karp is the founder and chief executive of Happiest Baby, Inc., which sells infant sleep products including a white-noise soother. Neither position has a trial behind it, so a reader should treat the question as open rather than settled.

That separate argument comes from animal work. A 2003 study in Science reared rat pups in continuous moderate-level noise and found that the normal maturation of the primary auditory cortex was delayed well beyond the usual developmental milestones, with the authors naming noise as a risk factor for abnormal development. Rats are not infants and a laboratory exposure is not a nursery, so this does not establish that a quiet sound machine affects a baby’s hearing or language. It is the reason the caution exists, and it is worth knowing that the reason is an animal study rather than a measurement in children. There is no trial that answers the question in humans, which is why the advice here stops at keeping the level low, the distance long and the duration no longer than it needs to be.

Dependence is a habit, and the evidence is thin

A common worry is becoming unable to sleep without the sound. What plausibly happens is a learned association: fall asleep to the same sound for months and a silent room becomes an unfamiliar sleep environment for a while. There is no body of research measuring this, so the honest answer is that it is unstudied rather than disproven — the confident claims found in either direction are not supported. If it concerns you, a fade-out timer or a gradual reduction costs nothing to try.

Worth separating from that worry: a sound that runs all night is also a sound that covers the noises you might want to hear, which is a practical cost rather than a health one. Nothing in the hearing literature speaks to it, and the fade and timer controls on this site exist partly because the end of a session is a choice a reader can make for themselves.

What the sleep evidence actually shows

The weakest part of the case for white noise is the part most often asserted. A 2021 systematic review in Sleep Medicine Reviews examined the literature on continuous noise as a sleep aid and graded the quality of evidence that it improves sleep as very low, noting that findings ran to both extremes, with some studies indicating that continuous noise disrupts sleep. The reviewers specifically remarked that this sits awkwardly beside how widely the practice is recommended.

That cuts both ways, and it is worth being precise about which way. There is no good evidence that normal-volume white noise is harmful to healthy adults, and no good evidence that it reliably improves sleep. It is a low-risk tool with an unproven benefit, which is a reasonable thing to try and a poor thing to rely on.

The separate question of whether the noise in a bedroom should be there at all has a much firmer answer. The WHO Environmental Noise Guidelines for the European Region review the evidence on environmental noise and health, including sleep disturbance, and set recommended exposure levels for road, rail, aircraft and wind-turbine noise. Reducing the intruding sound, with a window, a door or a different room, does something a masker cannot: it lowers the total exposure instead of adding to it. Masking is what is left when the source cannot be changed.

Using it without worrying about it

Simple guidelines for using white noise safely
SituationGuideline
VolumeStay well below 70 dBA, the level up to which long exposure is unlikely to harm hearing
All-night use (adults)Low volume, through a speaker rather than in-ear earbuds
Babies and nurseriesQuiet, device placed across the room, not at maximum and not longer than needed
Earbuds overnightAvoid; harder to judge the level and can irritate the ear canal
Headphone listening by dayThe WHO-ITU allowance is 40 hours a week at 80 dB, halving for every 3 dB more and doubling for every 3 dB less
Hearing or health concernsAsk a doctor or paediatrician about your situation

Used at a sensible level, white noise is not bad for you — the risk lives in the volume and the duration, which are both under your control, and in the distance between a sound machine and a crib, which is also under your control. Treat any reliance on it as a habit rather than a hazard, and treat the sleep benefit as plausible rather than established.

This article is general information, not medical advice. For concerns about hearing or your child’s health, consult a qualified professional.

References

Frequently Asked Questions

Is white noise bad for your brain?

There is no evidence that white noise at a normal volume harms the brain in healthy adults. It is a background sound used to mask distractions. The real risk with any audio played for hours is level and duration rather than the sound itself, so keeping it quiet is the precaution that matters.

Is white noise safe for babies?

It can be used to soothe infants, but with more care than for adults, and here there are measurements rather than opinions. A 2014 study in Pediatrics measured 14 infant sleep machines and found mean maximum output of 79 dBA at 30 cm, 71 dBA at 100 cm and 63 dBA at 200 cm; three exceeded 85 dBA at the closest distance, and all 14 exceeded the 50 dBA recommended for hospital nurseries at 30 cm. Place the device across the room rather than beside the crib, keep it quiet, and do not run it at maximum. Ask your paediatrician about your specific child.

Can you become dependent on white noise to sleep?

You can form a habit, where a silent room feels unfamiliar because you associate a particular sound with falling asleep. This is a learned association rather than harm. It is also essentially unstudied, so confident claims in either direction are not supported by evidence. If it bothers you, reduce use gradually or use a fade-out timer.

How loud is too loud for white noise?

The reference point is 70 A-weighted decibels: the NIDCD states that sounds at or below 70 dBA are unlikely to cause hearing loss even after long exposure, while long or repeated exposure at or above 85 dBA can cause it. Normal conversation sits at 60 to 70 dBA, so background sound you have to raise your voice over is louder than it needs to be. Lower is always safer, and masking does not depend on volume to work.

Is it bad to sleep with earbuds playing white noise?

No study has measured the hearing of people who sleep in headphones, so there is no documented harm to report, and the dose arithmetic is reassuring rather than alarming: the WHO-ITU allowance of 40 hours a week is 40 hours at 80 dB, and every 3 dB below that doubles the permitted time, so eight hours a night at a quiet background level is a small fraction of a week’s allowance. The argument for a speaker is a different one. An in-ear driver sits in the ear canal, which makes the level at the eardrum the hardest thing in this calculation to judge, and it can irritate the canal over eight hours. A speaker across the room is quieter at the ear and easier to estimate.

WhiteNoise.top

WhiteNoise.top builds free browser-based noise generators, mixers and breathing timers. Our articles explain the acoustics behind those tools and cite published research where it exists.