Does brown noise actually help you sleep?

Brown noise became a sleep aid the way most things become sleep aids now: a hashtag got big. By July 2022 #brownnoise had passed 62 million views on TikTok. What it has never had is a randomised controlled trial for sleep. Not a weak one — none at all.

Last updated September 2026 · about 7 minutes

That gap is worth sitting with, because almost everything written about brown noise and sleep is borrowed. The claims — it masks disturbances, it deepens sleep, it quiets a racing mind — are extrapolated from studies of white and pink noise, which are different signals, tested at different volumes, usually in a lab, usually for one or two nights. And as we'll see, that borrowed evidence is much shakier than the borrowing implies.

What brown noise actually is

The colours are a description of how sound energy is spread across frequencies, nothing more.

• White noise has equal power at every frequency. To the ear it is bright and hissy — untuned radio static.
• Pink noise drops off as frequency rises, by half the power per octave. It sounds fuller and more natural: steady rain, distant surf.
• Brown noise (also called red noise) drops off faster still. Most of the energy sits low down, so it reads as a deep rumble — a plane cabin, a heavy waterfall, a big fan two rooms away.

That's the whole distinction. Brown noise is not a different kind of thing from white noise, and there is no mechanism that would make it act on the brain in some category-separate way. It is the same broadband masking sound with the treble rolled off, which is mostly why people find it more comfortable to leave running all night.

The evidence that gets borrowed

The best summary of the field is a 2021 systematic review in Sleep Medicine Reviews by Riedy, Smith, Rocha and Basner, which gathered 38 studies of continuous noise used as a sleep aid. Its conclusion is the single most useful sentence anyone has written on this subject, and it is not the one you'll see quoted in app store listings: graded formally, the quality of evidence that continuous noise improves sleep is "very low" [1].

"Very low" is a technical rating, and it means something specific: the studies were small, often unblinded, inconsistent with each other, and measured different things in different ways. It does not mean noise has been proven useless. It means that if you set out to find out whether it works, you would have to start more or less from scratch.

Where it can go the other way

Two findings are worth knowing before you leave a masking sound running at volume all night.

Pink noise cost people REM sleep

In a study published in Sleep in 2026, a University of Pennsylvania group led by Basner tested pink noise and earplugs as ways to blunt intermittent environmental noise — traffic, aircraft, the neighbours. Twenty-five participants, measured with full polysomnography. The finding on the masking sound: 50 dB pink noise reduced REM sleep by roughly 19 minutes per night. Earplugs did better, mitigating nearly all of the environmental noise's effects without that cost [2].

Nineteen minutes is not a catastrophe, but it is a real trade. You are adding a constant sound to cover an intermittent one, and the constant sound is not free. If your actual problem is a noisy street, the boring answer — block the noise rather than paint over it — came out ahead in the one careful test we have.

The same sound helped one group and hurt another

A 2024 meta-analysis in the Journal of the American Academy of Child & Adolescent Psychiatry by Nigg and colleagues pooled 13 studies, 335 young people, on white and pink noise. In youth with ADHD it showed a small benefit. In non-ADHD controls the same noise impaired performance [3].

That was cognitive performance, not sleep, and children, not adults — so don't stretch it. But it is the clearest published demonstration of the thing that makes this whole question hard: a sound that helps the average person in one group can actively hurt the average person in another. Pooled results, applied to you individually, can point exactly the wrong way.

Volume is the part nobody mentions

A 2014 study in Pediatrics measured 14 infant sound machines. At 30 cm — roughly where one sits on a bassinet rail — all 14 exceeded the 50 dBA nursery limit, and three exceeded 85 dBA [6]. Eighty-five decibels is the level at which occupational exposure limits start applying to adults at work. For eight hours a night, next to an infant's ear.

The lesson generalises to adults with phones and speakers. If a sound is loud enough to mask everything, it is loud enough to be worth thinking about. Turn it down until it is just covering what you needed covered, and no further.

So should you use it?

Maybe. Here is what can be said without inventing anything:

• Brown noise masks. That is physics, not a clinical claim. If sudden noises wake you — a partner, a road, thin walls — a continuous sound raises the floor so those events stick out less.
• It might cue you. A consistent sound at bedtime becomes a signal, and signals are a real part of sleep habit. That is a claim about routine, not about the spectrum of the audio.
• The evidence for anything beyond that is very low quality, and at least one careful study found a cost in REM.
• If your problem is intermittent outside noise, earplugs beat pink noise in the one head-to-head trial available.

None of which tells you what happens in your bedroom. Group averages of very-low-quality studies are simply not the right instrument for a question about one person. The only instrument that is, is you — measured over enough nights.

How to find out for yourself

This is the problem SomniTune was built for. It plays the sounds, and then it does the part that's normally missing: it records which sound played at bedtime and lines it up against the night that followed — read from your Apple Watch, Oura, Whoop or Withings via Apple Health, or estimated from iPhone motion and on-device sound analysis if you don't sleep in a wearable.

After a couple of weeks you can see brown noise against rain against silence on your own nights, using your own falling-asleep time as the measure. It will not tell you what brown noise does to people. It will tell you what it does to you, which is the only version of the question you can actually act on.

Sources