Binaural Beats: Do They Work? What the Research Says

Cellular Coherence · Sound Lab

Binaural
Beats

Two slightly different tones, one in each ear, and your brain builds a third pulse that is not in the air. The pulse is well documented. Whether it reshapes brain waves, mood or sleep is where the evidence gets thin, so this page sorts what is established from what is hopeful.

Try it in the Sound Lab

How it works

A beat your brainstem builds

Mechanism first, claims second. Each step below carries an evidence tier, and the last one is where popular descriptions run ahead of the data.

Established Emerging Speculative or contested
01
Two tones, one per ear
Binaural beats occur when two tones of slightly different frequency are presented separately to each ear. The listener hears a third sensation pulsing at the difference between them. Example: 200 Hz in one ear and 210 Hz in the other gives a 10 Hz pulse.
Established
02
Built by the auditory brainstem
Reviews describe the signals converging in the superior olivary complex, with the inferior colliculus detecting the phase difference between ears. The beat is a product of neural processing, not a sound that exists in the air.
Established
03
Hard limits on what is heard
The effect was described by Oster in 1973. Reviews report it works mainly with low carrier tones below about 1000 Hz and is perceived for beat rates of roughly 1 to 30 Hz. Brainstem activity can follow the phase of incoming sound (the frequency-following response).
Established
04
Does the cortex follow along?
The brainwave entrainment hypothesis says the brain starts oscillating at the beat rate. A 2023 systematic review of 14 EEG studies found 5 supporting it, 8 contradicting it and 1 mixed. Claims that beats sync the two hemispheres or unlock states are not established.
Contested

Monaural beats, where the two tones are combined in the audio itself, have been tested too, but only one 2025 trial used them and it did not compare them with binaural beats. Isochronic tones are not covered here because I could not verify good comparative evidence. For the Lab tool and its brain-state selector, see Hemispheric Sync.

What the research shows

Small, mixed, low-certainty

Most of the trials are small, many are heterogeneous and the risk of bias is often high. Nothing here is a cure or a treatment. Effects below are statistical summaries from published reviews.

QuestionWhat was reviewedWhat it found, and caveatsEvidence tier
Perceived beatSystematic review of EEG studies, 2023Beats reliably create a perceived pulse; whether cortical brain waves entrain is inconsistent across 14 studies with very different methods. Contested
Anxiety, before surgeryMeta-analysis of 15 RCTs, 2025Lower anxiety than silence or blank audio (14 trials, 1,047 people, large effect), and still lower than non-binaural audio. Heterogeneity was extreme (I2 above 90%), so size is uncertain. Surgical patients are not everyday listeners. Emerging
Pain, around surgerySame 2025 meta-analysisLower postoperative pain vs blank audio (5 trials, moderate effect, high heterogeneity) and vs non-binaural audio (3 trials, I2 0%). Emerging
Pain and anxiety, dentalSystematic review of 9 studies, 2025Dental anxiety scores lower than controls; pain results inconclusive against no-intervention controls. Authors report high risk of selection and detection bias. Emerging
Pain, theta-range beatsMeta-analysis of 13 RCTs (630 people), 2026Pooled pain reduction from 4 trials (173 people, moderate effect) rated low-certainty. Anxiety and cognition were not pooled because outcomes differed. Authors cite few studies and low quality. Emerging
Memory, attention, anxiety, pain (pooled)Meta-analysis of 22 studies, 2019Overall medium effect (g = 0.45) across 35 effect sizes. Results depended on beat frequency, how long people listened and timing. Older and pooled across very different outcomes. Emerging
Young adults: anxiety, sleep, cognitionSystematic review of 10 trials, 2026Most trials showed small to moderate gains, but it mixes music-based and binaural interventions and the authors say heterogeneity and small samples limit certainty. Emerging
SleepNarrative review of music for sleep, 2025Included one binaural beat trial (43 people, two weeks) where theta beats did not significantly beat plain music on sleep quality. No firm sleep claim is supported. Speculative
MoodSingle RCT of monaural beats in music, 2025308 adults: beats embedded in music lowered anxiety and raised mood vs a pure tone; beats alone did not. Single track, industry-funded, monaural not binaural. Speculative
Plain verdict

Real pulse, unproven brain control

The pulse is real

You perceive a beat that the brainstem constructs. That part is solid.

Entrainment is contested

EEG evidence for cortical entrainment is inconsistent, so avoid any claim that a beat forces your brain into a state.

Benefits may be context, not tones

Comparisons matter. In one sleep trial theta beats did not significantly beat plain music, and in a 2025 trial beats inside music lowered anxiety while beats alone did not.

Evidence is low certainty

Small samples, heterogeneity, high bias risk and some industry funding. Worth experimenting with, not worth betting health decisions on.

Safety

How to use them safely

Seizure caution

If you have epilepsy or a seizure history, or are unsure, talk to your clinician before using beats. I did not find controlled safety data either way, so caution is the sensible default.

Headphones, kept quiet

Binaural beats need stereo headphones so each ear gets its own tone. Keep volume comfortable and low, as a 2025 sleep review also advises.

Stop if it feels wrong

Stop if you get dizziness, headache, ear discomfort or agitation. Do not listen while driving or doing anything that needs full attention.

Not a treatment

This page is educational, not medical advice. Beats do not cure or treat any condition. For anxiety, pain or sleep problems, see a qualified professional.

Practical notes

What the studies suggest

01
Pick low carriers
Reviews say the effect works mainly with carrier tones under about 1000 Hz and beat rates around 1 to 30 Hz.
Established
02
Listen for a while
In the 2019 meta-analysis, longer exposure and listening before a task did better than only during it. Several trials use around 30 minutes. No optimal dose is verified.
Emerging
03
Skip the masking noise
The same meta-analysis found unmasked beats worked similarly to beats masked with white or pink noise.
Emerging
04
Treat it as an experiment
Notice your own response, compare against plain calm audio, and keep what helps.
Your own trial
Try it

Test it yourself in the Sound Lab

Open the player with a binaural recipe, use headphones at low volume and judge the effect against plain quiet. See also Hemispheric Sync and Frequency Healing.

Sources

Where this comes from

  1. Xiong et al. 2025, perioperative anxiety and pain meta-analysis (PubMed)
  2. Shukla et al. 2025, dental pain and anxiety systematic review (PubMed)
  3. Garcia-Argibay et al. 2019, cognition, anxiety and pain meta-analysis (PubMed)
  4. Ingendoh et al. 2023, brain oscillation systematic review, PLOS ONE (PubMed)
  5. Elnazer 2026, music and binaural beats in young adults (PubMed)
  6. Theta binaural beat therapy meta-analysis, EXPLORE 2026 (ScienceDirect abstract)
  7. Pan and Wang 2025, music elements for sleep, narrative review (Frontiers in Sleep)
  8. Mirmohamadi et al. 2024, A Review of Binaural Beats and the Brain (PMC)
  9. Venkatesan et al. 2025, monaural beats randomized trial (Frontiers in Psychology)
Educational content, not medical advice. Last fact-checked November 2026. Summaries are paraphrased from the abstracts and pages listed above.
Sound Lab
Scroll to Top