What do the studies actually show about binaural beats?

What are binaural beats?

Play a tone at 200 Hz in one ear and 210 Hz in the other, and you hear a pulse at 10 Hz that exists nowhere in the room. Your auditory system creates it from the difference. Binaural beats are that illusory pulse, and headphones are required because each ear needs its own tone.

The claim built on top of this is where things get interesting. The popular account says the brain synchronises its own electrical rhythms to that phantom pulse, so a 10 Hz beat pulls you toward a relaxed alpha state and a 40 Hz beat toward alert gamma. This idea is called brainwave entrainment, and almost every product in this category rests on it.

Worth separating two questions from the start. Whether listening does anything useful is one question. Whether it works by entraining your brainwaves is a different one. The research answers them differently.

Do binaural beats actually work?

The pooled evidence points to a modest real effect. A meta-analysis of 22 studies covering memory, attention, anxiety and pain reported an overall effect size of g = 0.45, described by its authors as medium and consistent. Individual studies vary a great deal, and the effect depends on frequency, timing and duration.

That meta-analysis, by Garcia-Argibay and colleagues, is the most cited quantitative summary in the field. It is named here in prose because the record could not be opened directly for checking. Its practical findings are worth carrying: exposure before a task, or before and during it, outperformed exposure during the task alone, longer listening periods did better than short ones, and masking the beats with white or pink noise made no difference to whether they worked.

A 2025 systematic review and meta-analysis in the International Journal of Clinical Pediatric Dentistry looked specifically at dental pain and anxiety across nine studies. It found meaningful reductions in anxiety against control, and a more tangled picture for pain, where the comparison against no intervention was inconclusive with very high heterogeneity. The authors graded the overall certainty as low and flagged a high risk of selection and detection bias.

Taken together: something is happening, it is probably real, and it is smaller and less reliable than the marketing around it suggests. The anxiety findings are the most consistent part of the picture.

Does the brainwave entrainment explanation hold up?

Weakly. A 2023 systematic review in PLoS One examined 14 studies measuring brain activity during binaural beat exposure. Five supported the entrainment hypothesis, eight contradicted it, and one was mixed. No study using beats in the beta range found entrainment at all.

Ingendoh, Posny and Heine, who conducted that review, concluded that the results corroborate an overall inconsistency of empirical outcomes. Where positive findings did appear, they showed up as enhanced EEG power in theta, alpha and gamma bands. The picture across studies was too heterogeneous to pool statistically, so they could only describe it.

This matters for how you read every product claim in this category. A page telling you that a 40 Hz track puts your brain into gamma is stating as fact something the measurement literature splits roughly two to one against.

A more cautious reading of the useful effects is that sustained, predictable, non-distracting sound supports attention and lowers arousal, and the beat frequency may matter less than the listening itself. That account fits the evidence at least as well and promises you less.

Does brain.fm work?

Brain.fm has better evidence behind it than binaural beats generally, and it is worth knowing that it does not use binaural beats. Its technology is amplitude modulation, a rapid pulsing of the music volume. A 2024 paper in Communications Biology found it improved sustained attention, most clearly in listeners with higher ADHD symptom scores.

The study, led by Woods with colleagues including Psyche Loui at Northeastern, ran four experiments: 83 people on a behavioural attention task, 34 in fMRI, 40 in EEG, and 175 in a task varying modulation rate and depth. Rapidly modulated music improved performance on a sustained attention task, activated salience and executive control networks in fMRI, and produced stronger phase-locking between the sound and brain activity in EEG. The clearest benefit appeared at 16 Hz modulation among participants scoring higher for ADHD symptoms.

The conflict of interest is declared in the paper and belongs in any honest answer. Kevin Woods, the lead author, is an employee of Brain.fm and holds unexercised equity options in the company, which produced the music used in the study. A second author was previously employed there with equity. The work was peer reviewed, published in a Nature portfolio journal, partly funded by the US National Science Foundation, and conducted with an academic laboratory. Company involvement of this kind is common in product research and it is a reason for measured confidence rather than dismissal. Independent replication by a team with no financial stake is what would settle it.

So the fair answer to whether brain.fm works: there is real, peer-reviewed evidence for the specific mechanism it uses, stronger than what supports binaural beats, and it comes from a study its own employee led. If you have attentional difficulties, it is among the better-supported options in this category. If you are looking for a proven general intelligence or productivity boost, no product in this space has that.

Endel, which generates adaptive soundscapes, has considerably less published evidence behind it. The honest position on Endel today is that the underlying idea of steady non-distracting sound is reasonable, and the specific product claims are not backed by the kind of independent trial evidence that would let anyone say it works.

What should you actually do with this?

If sound helps you focus or settle, keep using it. The evidence supports steady, predictable, non-distracting audio more strongly than it supports any particular frequency. Start before the task rather than partway through, give it longer than a few minutes, and judge it by whether your work improves.

Two practical points come straight out of the meta-analysis. Beginning to listen before you start working produced better results than starting once you were already going, and longer exposures beat brief ones. Both are easy to apply.

Headphones are genuinely required for binaural beats, since the effect depends on each ear receiving a different tone. For amplitude-modulated music of the brain.fm kind, and for plain steady sound like brown noise or rain, speakers are fine.

The most reliable test is your own work. Sound that helps will show up as more time on task and less pull toward your phone. Sound that does nothing will feel pleasant and change nothing measurable, which is worth knowing before you pay a subscription for it.