White noise for sleep: what the research does (and doesn't) show

Does white noise help a child fall asleep?

Possibly, but it isn't proven. The most cited study on the topic, published in 1990, found an effect of white noise on newborns falling asleep. Spencer, Moran, Lee, and Talbert randomly split forty newborns aged two to seven days into two groups of twenty, one exposed to continuous white noise, the other not. Sixteen of the twenty newborns in the white noise group fell asleep in under five minutes, compared with five of twenty in the control group.

That gap has since fed the idea that white noise speeds up falling asleep, a plausible hypothesis, since continuous noise can mask the small household sounds that would otherwise catch an infant's attention. The study is limited, though, to newborns just a few days old, measured over a single episode of falling asleep: it says nothing about the effect of white noise over several nights, or in older children.

What other sources call into question

In 2021, Riedy, Smith, Rocha, and Basner reviewed the published studies on continuous noise as a sleep aid: thirty-eight in total. Their finding has less to do with whether an effect exists than with how solid the evidence for it is. The noises played, the sleep measurement methods, adherence to the intervention, and the control conditions differed greatly from one study to the next.

Perhaps partly as a result, the findings range from noise that improves sleep to noise that disrupts it. Under the GRADE criteria they applied, the quality of the evidence for claiming that continuous noise improves sleep is very low, which, they write, contradicts its widespread use. They add that continuous noise may also have a negative effect on sleep and hearing.

Why the disagreement persists

Studies on white noise don't all measure the same thing in the same way: the type of noise played, how long the exposure lasted, the children's ages, how sleep was defined (parent report, observation, or recording), and whether or not there was a control group. Each of these variables changes from one study to the next, which makes the results hard to compare, and even harder to pull together into a single conclusion.

The authors of the 2021 review call for research with objective sleep measures and detailed descriptions of noise exposure before continuous noise is promoted as a sleep aid. The scientific disagreement here reflects less a set of contradictory results than a lack of data solid enough to settle the question either way.

How loud should a white noise machine be, and how far from the bed?

As far from the bed as possible, at the lowest volume. On sound exposure itself, the available measurements are clear. Hugh, Wolter, Propst, and colleagues published their study in Pediatrics, the American Academy of Pediatrics' journal. They tested fourteen infant white noise machines at maximum volume, at 30, 100, and 200 cm from the source, with corrections for the ear of a 6-month-old. On average, these machines reached 79 dB at 30 cm, 71 dB at 100 cm, and 63 dB at 200 cm.

The benchmark used in hospital nurseries is 50 dB: every machine exceeded it at 30 cm, and almost all still exceeded it 200 cm from the crib. Three of them exceeded 85 dB, a level that, kept up for more than eight hours, goes beyond the occupational exposure limits for an adult.

These measurements were taken at maximum volume: they don't say what level you reach at a usual setting, but they show that the setting matters. The authors draw simple rules from them, whichever machine you choose: place it as far as possible from the crib, never inside it or on its rail; set the volume as low as possible; and limit how long it runs rather than leaving it going all night at full volume.

How to decide without relying on a marketing claim

An effectiveness figure printed on a machine's packaging is worth checking: the 2021 review concludes that the quality of the evidence is very low, which doesn't allow anyone to promise an effect. What the existing research shows is enough to set reasonable expectations: a possible effect on falling asleep, never proven rigorously enough to be guaranteed.

Trying white noise for a few nights, at low volume and at a safe distance from the bed, tells you more about its usefulness for a particular child than any product page. For example: the machine sits on a shelf at the far end of the room, not on the crib rail, at the lowest volume, with a one-hour timer. If it produces nothing noticeable after one to two weeks, stopping it costs the child no proven benefit; if it helps, the volume and distance rules above remain the only ones that matter, regardless of the debate over its effectiveness.

For temperature, light, and the rest of the bedroom, see the guide on the bedroom environment. If the baby sleeps poorly despite everything, or if there is any doubt about their hearing, the pediatrician is the right person to ask.