Binaural Beats — Body Handbook

Psychology · §447

Binaural Beats

Binaural beats are two pure tones — one piped into each ear through headphones — whose tiny frequency difference your brain hears as a third, pulsing "beat" that nobody else can record. Apps sell them as a brainwave entrainment tool: play a 4 Hz beat, the marketing says, and your whole brain slides into a meditative state. Careful brain-recording studies say no — the brainstem does pick up the beat, but the global brain shift the marketing promises mostly doesn't show up. What does show up is a real, replicated drop in anxiety, especially before stressful events. Everything else — focus, sleep, deep meditation — is small, inconsistent, and often indistinguishable from listening to any calming audio.

Know · As-needed Evidence Mixed Chapter Psychology

The honest pitch: free, ten minutes on headphones, a real but modest lever against anxiety. Reach for it before a presentation, a job interview, a dentist appointment, and you're likely to come out calmer — about the same lift you'd get from breathwork, and not much beyond what pleasant music alone delivers. Reach for it expecting transformative focus, deeper meditation, or instant sleep, and you'll be disappointed. The brainwave-entrainment story most apps lean on is the part of the claim the literature has quietly walked back.

The percept itself is real and uncontroversial. Two tones close in pitch — 440 Hz in your left ear, 444 Hz in your right — can't fake the beat from a single speaker; the integration that produces the pulsation happens inside your head, in the brainstem, where signals from the two ears first meet. Listen on headphones and you hear a steady 4 Hz throb that doesn't exist in either ear's signal alone. Take one headphone off and the beat disappears. Oster 1973 described this in Scientific American fifty years ago, and nobody disputes it now.

The contested step is what happens above the brainstem. The marketing claim is "brainwave entrainment": the beat at 4 Hz pulls your cortex into 4 Hz oscillation, and you get whatever mental state that band is associated with — theta for meditation, alpha for relaxation, beta for focus, delta for sleep. The auditory cortex does pick up the beat as a steady, beat-matched signal. The rest of the brain doesn't reliably follow. A 2020 study compared binaural beats against plain monaural beats — one ear, same difference frequency — and found that the simpler monaural version actually produced stronger cortical entrainment Orozco Perez et al. 2020. On the EEG evidence, the "two ears = special phenomenon" framing the marketing leans on is the wrong way round.

What the trials actually show

The headline finding is anxiety reduction in the moments before a stressful event. People going under for surgery, sitting in a dentist's chair, or about to dissect a cadaver in anatomy class consistently report lower anxiety after a short pre-event listen than matched controls without the beats. The effect is real, it replicates, and it's large enough to feel.

Look closer and the picture is messier. The 2024 review that synthesised the non-clinical, "I just listen at home" stress-management trials flagged how heterogeneous the field is — different beat frequencies, different durations, different masking sounds, different outcome measures, and almost no reporting of side effects at all Platt & Hammond 2024. A separate 2024 review focused on anxiety and depression came out broadly positive: beats — pure or layered onto other audio — outperform silence and noise-cancelled controls Baseanu et al. 2024. The signal is there. It's just much smaller than the consumer marketing implies, and it shrinks when the blinding gets stricter.

Sleep is where the evidence thins out fastest. The cleanest study — a double-blind sham-controlled trial in 43 adults with mild ongoing insomnia — found no benefit from theta beats over plain music after two weeks Bang et al. 2019. Smaller unblinded pilots have reported shorter sleep-onset latencies from low-frequency delta beats Dabiri et al. 2022, but the moment a credible placebo enters the design the effect tends to drop out.

Focus and working memory are similarly inconsistent. A 2023 meta-analysis specifically targeting attention and memory described the literature as potential but inconsistent — some trials show clear improvement, some show nothing, and a few show worse performance under the beats than under control conditions Basu & Banerjee 2023. There is no reliable evidence base for the deep-meditation effect at all; the trial literature simply hasn't tried to measure meditative depth in a way that would tell you whether binaural beats produced it.

What the apps oversell

The marketing story almost every product leans on goes like this: your brain has a natural rhythm for each mental state — delta for sleep, theta for meditation, alpha for calm, beta for focus — and playing a beat at one of those frequencies pulls your whole brain into that state. It's a clean, sellable story. It's also not what brain recordings show. A 2023 systematic review pulled every careful EEG study testing the idea — 14 in total — and found that 8 contradicted whole-brain entrainment, 5 supported it, and 1 was mixed Ingendoh et al. 2023. Where entrainment does happen, it's mostly in auditory cortex — the part of the brain that processes the sound itself — not the prefrontal or limbic regions whose activity would actually produce the meditative or focused state being advertised.

A related myth: that Gerald Oster invented brainwave entrainment with binaural beats in his 1973 Scientific American article. He didn't. Oster was interested in binaural beats as a way to detect auditory processing differences and possibly neurological conditions like Parkinson's Oster 1973. He never claimed they entrain brainwaves at all; that framing arrived later, mostly from the consumer side, and now travels under his name as borrowed credibility.

One more: the idea that you need a special app or a high-bitrate file. The brainstem doesn't care. As long as the two tones reach two different ears at roughly matched volume, the beat shows up. A free YouTube track played through cheap stereo headphones does what a $15-per-month app does, modulo the ads and the interface.

If you want to try it

The realistic use case is pre-event anxiety reduction. You have a presentation in an hour, you have to call your dentist, you're about to walk into something that's going to spike your heart rate. Put on the headphones, sit somewhere reasonable, and listen for 10–20 minutes before the thing — not during it. Pre-task listening consistently outperforms during-task listening in meta-analysis Garcia-Argibay et al. 2019.

Where to find tracks: any streaming service, any YouTube search for "binaural beats" plus the band you want, or a dedicated app (Brain.fm, Endel, BinauralPure) if you want curated content with a built-in timer. The audio itself is free; apps charge for convenience and interface.

What you'd otherwise reach for

For pre-event anxiety, the better-evidenced alternatives are usually larger and faster. Three minutes of slow nasal breathing — 4 seconds in, 7 seconds hold, 8 seconds out — drops heart rate and subjective anxiety reliably. Box breathing (4-4-4-4) does similar work. A short mindfulness pause, progressive muscle relaxation, or a brisk five-minute walk are all in the same low-effort, real-effect bucket. For sleep onset, the bigger levers are consistent sleep timing, light exposure during the day, and — when the problem is persistent — cognitive behavioural therapy for insomnia. For meditation depth, the established traditions and guided-practice apps have decades of trial evidence behind them; binaural beats don't.

None of that means binaural beats are pointless. It means the right mental category for them is "another option in the calming-audio toolbox" — alongside ambient music, nature sounds, pink noise — rather than a separate-class technology that does something the others can't.

Why it didn't work for you

The most common reasons a real attempt produces nothing:

Speakers, or a single earbud. No two-ear separation, no beat percept. Half the people who say "I tried binaural beats and felt nothing" tried them on a phone speaker.
Listening during the task rather than before it. Pre-task exposure beats during-task in meta-analysis Garcia-Argibay et al. 2019. A 15-minute listen, then start the work.
A track where the beat is barely audible under the music. You can't always tell from the title. If the pulsation isn't clearly perceptible, the beat layer is decorative and you're listening to ambient music — which has its own (smaller) effect, but isn't doing the thing the beats are meant to do.
Expecting a transformation. The trial-grade effect sizes are small to moderate. People primed by Reddit threads to expect a life-changing focus state read the actual modest lift as failure.
Genuine individual variability. Some listeners can't perceive binaural beats at all; perceptibility varies between people and within the same person across time. Oster noted this in 1973 Oster 1973.
Daily use for months expecting growing effect. The trial literature is concentrated in 1–8 week protocols. Tolerance and habituation past that point are barely studied; assume diminishing returns.

Related routes worth knowing about: isochronic tones and monaural beats, closely related entrainment audio that doesn't need headphones and in some studies actually produces stronger cortical effects than the binaural version; audio-visual entrainment headsets that combine flashing light with sound; transcranial alternating-current stimulation (tACS), a research technique that genuinely does drive cortical oscillations electrically but is invasive; and the broader sound-therapy and meditation traditions, where the entrainment claim takes a different form and rests on different evidence.

Related in the handbook

Alternatives

— Another low-effort calming tool, though autogenic training builds a deeper effect over weeks.
— For pre-event nerves, slow breathing gets you about the same calm — and needs no headphones.
— Apps sell beats as instant meditation; the real thing does more, it just takes weeks of practice.

— Both are press-play audio for calm — if the beats don't land, a guided NSDR track is another low-effort way to drop the stress dial.

Substance + claimed effects

Binaural beats are a perceptual phenomenon: two pure tones of slightly different frequency are presented dichotically (one to each ear, through headphones), and the listener perceives a phantom third tone — the "beat" — oscillating at the difference frequency between the two carriers. The phenomenon was first described by the Prussian physicist Heinrich Wilhelm Dove in 1839 and reintroduced to modern audiences by biophysicist Gerald Oster in his 1973 Scientific American article (Oster 1973). Typical commercial implementations use carrier tones in the 200–900 Hz range with beat frequencies of 1–30 Hz, mapped onto the canonical EEG bands: delta (<4 Hz), theta (4–8 Hz), alpha (8–13 Hz), beta (13–30 Hz), gamma (>30 Hz). The central claim, popularised by consumer products from the 1980s onward (Monroe Institute "Hemi-Sync", Holosync, countless YouTube channels and apps), is brainwave entrainment: that exposing the listener to a 4 Hz beat will pull cortical oscillations toward 4 Hz, producing whatever mental state is associated with that band — relaxation, focus, sleep onset, meditative depth, anxiolysis. This entry covers all five effects named in the brief (focus, relaxation, sleep onset, anxiety reduction, meditative states) plus the supporting evidence quality, which is the load-bearing question — the gap between what consumer marketing claims and what the literature actually supports is the single most important fact a reader can leave with.

Evidence by addressing question

Mechanism

The perceptual mechanism is well established and uncontroversial at the brainstem level. When the two ears receive tones that differ by less than ~30 Hz, the superior olivary complex in the brainstem integrates the inputs and generates a frequency-following response (FFR) at the difference frequency (Oster 1973; Orozco Perez et al. 2020). This is a real, measurable neural signal — the brainstem genuinely tracks the beat. Oster called the perceptual window the "critical bandwidth": carriers must be within ~18% of each other for the beat to be heard, and the effect requires dichotic presentation (a speaker in each room won't do; the integration is neural, not acoustic).

The contested step is whether this brainstem FFR propagates upward into broad cortical entrainment — i.e., whether a 10 Hz beat actually pulls alpha-band activity across sensory and prefrontal cortex into phase with it, the way photic stimulation can in some preparations. Orozco Perez et al. used EEG with source localisation and explicit comparison to a monaural-beat control. They found a subcortical FFR at the carrier and an auditory steady-state response (ASSR) at the beat frequency in auditory cortex, but the cortical entrainment was weaker than that produced by ordinary monaural beats — the binaural version's advertised advantage was empirically reversed (Orozco Perez et al. 2020). The early Wahbeh et al. pilot (n=4) similarly failed to find steady-state EEG entrainment to a 7 Hz binaural beat (Wahbeh et al. 2007). Mechanism is therefore partial: brainstem tracking is real, primary auditory cortex picks up the beat as an ASSR, but the "whole-brain shift into target band" the consumer marketing implies is not established.

Evidence

The literature has three meta-analytic landmarks. Garcia-Argibay et al. (2019) pooled 22 studies / 35 effect sizes on cognition, anxiety, and pain perception and reported an overall Hedges' g = 0.45 — a moderate effect size, statistically significant, with the largest signal in the theta/delta range for anxiety (g = 0.69 across 5 effect sizes / 4 studies / n = 159) (Garcia-Argibay et al. 2019). They also found that pre-task exposure outperformed during-task exposure and that pure beats outperformed beats embedded in music. Basu & Banerjee (2023) ran a more recent meta-analysis focused on memory and attention; they reported potential improvements but flagged inconsistent results across studies, with some trials showing improvement, some null, and some worsened performance (Basu & Banerjee 2023). Ingendoh et al. (2023) performed the most relevant systematic review for the brainwave-entrainment claim itself: of 14 EEG studies meeting quality criteria, 5 supported entrainment, 8 contradicted it, and 1 was mixed; methodological heterogeneity was severe (Ingendoh et al. 2023).

The single most replicated behavioural finding is anxiety reduction in clinical/perioperative settings. Padmanabhan et al.'s prospective RCT in 108 day-case surgery patients reported a 26.3% drop in State-Trait Anxiety Inventory score in the binaural group vs. 11.1% with matched music and 3.6% with no audio (p = 0.001) (Padmanabhan et al. 2005). Multiple subsequent perioperative trials (cataract, pterygium, cadaveric dissection) replicate the direction; a recent perioperative meta-analysis reported SMD = −1.01 for anxiety vs. non-binaural audio (a large effect, with wide CI). The 2024 Platt & Hammond systematic review of non-clinical, personal-use stress management found mixed-to-positive effects across heterogeneous RCTs and flagged that adverse-event reporting in this literature is essentially absent (Platt & Hammond 2024). The 2024 Baseanu et al. review of anxiety and depression outcomes concluded that beats — pure or masked — outperform no-music or noise-cancelling controls (Baseanu et al. 2024).

Sleep is much weaker. Bang et al.'s double-blind sham-controlled RCT in 43 subclinical-insomnia adults found that theta binaural beats embedded in music produced no significant improvement in insomnia severity beyond music alone (p = 0.656) — the headline finding was "minimal effects" (Bang et al. 2019). The Dabiri et al. 2022 pilot in 20 healthy students reported shortened sleep-onset latency and improved subjective sleep quality after a week of 3 Hz delta beats — but the design was pre-experimental with no sham control, so placebo and routine effects are uncontrolled (Dabiri et al. 2022). Focus and working-memory trials are likewise small, heterogeneous, and often null in sham-controlled designs.

Protocol

Where the literature reports a protocol, it converges on roughly: stereo headphones, carrier tones in the 200–500 Hz range, beat frequency in the target EEG band (commonly 4 Hz for relaxation/meditation, 10 Hz for anxiolysis and "alpha relaxation", 14–18 Hz for arousal/attention, <1 Hz for sleep), 10–30 minutes of listening, applied before or before-and-during the target task (pre-task exposure being meta-analytically superior to during-task) (Garcia-Argibay et al. 2019). Perioperative trials typically used 10 Hz alpha beats for ~10–15 minutes immediately before induction (Padmanabhan et al. 2005). Wahbeh et al.'s delta-frequency trial used 30-minute sessions over 60 days for chronic anxiety (Wahbeh et al. 2007). Carrier-tone amplitude needs to be loud enough to hear comfortably and identical in both ears; volume modulation across ears destroys the beat. Tracks embedded in music produce smaller effects in meta-analysis than pure beats (Garcia-Argibay et al. 2019), though listeners tolerate music-embedded tracks longer.

Contraindications

The auditory-entrainment safety question is meaningfully different from photic entrainment. Strobe-light stimulation reliably provokes seizures in photosensitive epilepsy; no equivalent literature establishes that pure auditory entrainment at sub-30 Hz beat frequencies provokes seizures in the general population. The Platt & Hammond 2024 review flagged that adverse-event reporting in binaural-beat RCTs is "essentially absent" and called for CONSORT Harms 2022-compliant reporting in future trials (Platt & Hammond 2024). There is no clear empirical signal of seizure provocation in the published literature, but the conservative clinical position remains that individuals with diagnosed epilepsy — particularly musicogenic, pattern-sensitive, or reflex epilepsies — should consult a neurologist before structured use. A second flag: Wahbeh et al.'s pilot reported increased depression scores and poorer immediate recall in some participants (Wahbeh et al. 2007); n was small (4), but it is the only adverse-direction effect with controlled measurement, and Garcia-Argibay et al. note that some studies show worsened cognitive performance (Garcia-Argibay et al. 2019). Headphone listening at unsafe volumes carries the standard hearing-damage risk independent of binaural content. Pregnancy, cardiac conditions, and the other catalogue-vocabulary contraindications do not apply.

Misconceptions

The dominant misconception is the entrainment-equals-state-shift claim: that listening to a 4 Hz beat "puts you in" a theta state with all its associated meditative or creative properties. The Ingendoh et al. 2023 review establishes that whole-brain entrainment at the beat frequency is not consistently observed in EEG, with more studies contradicting it than supporting it (Ingendoh et al. 2023); Orozco Perez et al. specifically showed that the cortical entrainment binaural beats do produce is weaker than what monaural beats (a single ear, simpler stimulus) produce (Orozco Perez et al. 2020). A second misconception, popular in product marketing, attributes binaural-beat brainwave entrainment to Oster — but Oster never claimed it; he was interested in binaural beats as a diagnostic tool for auditory processing differences and neurological conditions (Oster 1973). A third misconception: that beat-frequency selection is causal. Garcia-Argibay et al. find effects across multiple beat-frequency bands and note that musical and emotional context often matters as much as the chosen frequency (Garcia-Argibay et al. 2019). A fourth: that ear-bud quality matters. Cheap stereo headphones reproduce 200–500 Hz tones faithfully; the beat is generated in the brainstem, not the driver.

Alternatives

For anxiety reduction, the immediate alternatives are calming music (which produced about a third of Padmanabhan's anxiety reduction in the same trial), guided breathwork (4-7-8, box breathing, slow nasal breathing), brief mindfulness, and progressive muscle relaxation — all with comparable or larger effect sizes in their respective literatures. The meta-analytic effect of "music with binaural beats" vs. "music alone" is often modest, suggesting the beats add incrementally to a base effect produced by the audio context itself (Bang et al. 2019; Baseanu et al. 2024). For sleep onset, CBT-I, sleep hygiene, and (where clinically indicated) melatonin or sedating antidepressants have substantially larger evidence bases. For meditative states, established meditation traditions and apps with guided practice are the higher-confidence routes. Binaural beats' positioning is therefore as a low-friction, free, headphone-based add-on, not a primary intervention.

Failure modes

Common reasons binaural beats "didn't work" for a given user: (1) speakers instead of stereo headphones — no dichotic presentation, no beat, no FFR; (2) listening during the task rather than before it — pre-task exposure is meta-analytically superior (Garcia-Argibay et al. 2019); (3) brief one-off use against an expectation set by Reddit anecdotes of "transformative" effects, when the actual effect sizes are small-to-moderate; (4) the chosen track is mostly music with minimal beat amplitude, so the beat is effectively decorative; (5) individual variability is substantial — some listeners cannot perceive the beat at all, and Garcia-Argibay et al. note person-level variance in the effect (Garcia-Argibay et al. 2019); (6) expectation effects: studies that compared belief-induced expectancy vs. neutral framing show that expectancy alters outcomes, complicating any "real" effect attribution. Some studies actually report worsened cognitive performance, possibly because a fixed external frequency interferes with the listener's natural rhythm rather than reinforcing it (Garcia-Argibay et al. 2019; Basu & Banerjee 2023).

Practicalities

The practicality story is the substance's main competitive advantage: free or near-free (YouTube, Spotify, dedicated apps), zero learning curve, zero physical risk in non-clinical populations, and a 10–30 minute time cost. Any consumer headphones reproducing 200–500 Hz cleanly suffice — no special hardware. Apps like Brain.fm, Endel, and BinauralPure package the experience with curated tracks and timers; free tracks on YouTube are functionally equivalent if the user can ignore advertising. The Wahbeh et al. clinical protocol was 30 minutes per day for 60 days for chronic anxiety (Wahbeh et al. 2007); most consumer use is 10–20 minutes on demand.

History

Dove described the binaural-beat percept in 1839 in a Berlin physics report. The phenomenon sat as a curiosity in psychoacoustics for a century. Gerald Oster's 1973 Scientific American article (Oster 1973) reframed it as a tool for investigating auditory neurology and proposed diagnostic uses (Parkinson's, hormonal cycles, hearing-pathway disorders). The Monroe Institute commercialised "Hemi-Sync" tapes in the late 1970s, framing binaural beats as a consciousness-altering technology — a framing that has dominated consumer marketing ever since. The first peer-reviewed clinical trial of binaural beats for anxiety appeared in 2001 (Le Scouarnec); the field expanded steadily after 2005 with Padmanabhan and Wahbeh, then accelerated post-2015 with the proliferation of streaming-audio platforms.

Out of scope

Isochronic tones and monaural beats (closely related entrainment audio that does not require dichotic presentation); audio-visual entrainment / brainwave-entrainment headsets (Mindplace, Roxiva — different modality, different evidence base); transcranial alternating-current stimulation (tACS), which actually does drive cortical oscillations electrically but is invasive; broader sound-therapy traditions (Tibetan bowls, gong baths, drumming) where the entrainment claim differs in form; clinical neurofeedback (closed-loop EEG training).

Credibility range

Optimist case. Brainstem FFR at the beat frequency is real and uncontroversial; auditory cortex generates a measurable ASSR at the beat frequency (Orozco Perez et al. 2020). Meta-analysis across 22 trials gives a moderate effect size for cognition, anxiety, and pain perception (g = 0.45), with the strongest signal in theta/delta-frequency beats for anxiety (g = 0.69) (Garcia-Argibay et al. 2019). The single most replicated clinical signal — perioperative anxiety reduction — is robust across multiple independent RCTs and produces effect sizes large enough to matter clinically (Padmanabhan et al. 2005). The intervention is free, near-zero risk, headphone-based, and 10–30 minutes — a remarkably favourable cost-benefit profile even if effects are modest. Community-level reports (millions of users, decades of consistent thematic reports of relaxation and focus support) are consistent in direction even where they don't reach trial-grade rigour. Adverse-event signal in the literature is essentially absent (Platt & Hammond 2024). For a low-stakes complementary practice, the evidence-to-effort ratio favours trying it.

Skeptic case. The central mechanistic claim — that binaural-beat exposure entrains whole-brain cortical oscillations into the target band, the way the marketing implies — is not supported. Ingendoh et al.'s systematic review finds more EEG studies contradicting brainwave entrainment than supporting it (8 vs 5 vs 1 mixed) (Ingendoh et al. 2023). Orozco Perez et al. specifically show that simpler monaural beats produce stronger cortical entrainment than binaural beats, undermining the "two ears = special phenomenon" framing (Orozco Perez et al. 2020). Behavioural effect sizes shrink dramatically in sham-controlled designs: Bang et al.'s double-blind sham-controlled RCT found no benefit beyond music alone for subclinical insomnia (Bang et al. 2019); much of the field is small, single-site, unmasked, with high methodological heterogeneity (Basu & Banerjee 2023; Ingendoh et al. 2023). The perioperative anxiety signal is plausibly attributable to general audio-distraction plus expectancy, since matched music produces a substantial fraction of the same effect in the same trials. Adverse-event reporting being absent is not evidence of safety; it's evidence of lazy reporting (Platt & Hammond 2024). Commercial actors — Monroe Institute, Holosync, dozens of YouTube channels with millions of subscribers — have strong incentives to overstate.

Author's call. The substance produces a real-but-modest anxiety-reduction effect, well-replicated in perioperative settings, plausibly extending to non-clinical stress contexts; this is the single defensible reader-facing claim. Effects on focus, sleep onset, and meditative depth are smaller, less consistent, and frequently null in sham-controlled designs. The "brainwave entrainment" mechanism as marketed — pulling whole-brain activity into the target frequency band — is not supported by current EEG evidence; what is supported is brainstem FFR and an auditory-cortex ASSR, neither of which obviously translates to the global cognitive-state shifts consumers are promised. The honest framing is: binaural beats are a free, low-risk, headphone-based add-on for relaxation and pre-task anxiety reduction, working through mechanisms that probably include distraction, expectancy, calming audio, and some genuine but limited auditory entrainment — not the brainwave-state alchemy the marketing suggests. Evidence rating 2, controversy rating 3: sparse and contested literature; mechanism plausible at brainstem level but the consumer-facing claim runs ahead of the data; active disagreement among reasonable researchers about whether the cortical entrainment is real and clinically meaningful.

Stakeholder + incentive map

Commercial: Monroe Institute (Hemi-Sync, the historical incumbent), Holosync (Centerpointe), Brain.fm, Endel, BinauralPure, and a long tail of app and YouTube monetisation. Strong financial incentive to overstate; subscription products often pitch month-long protocols framed as treatments.
Academic: A small but growing cluster of psychoacoustics and clinical-anaesthesia researchers (Padmanabhan, Wahbeh, Garcia-Argibay, Orozco Perez, Ingendoh, Bang). Mixed incentive: positive trials are publishable, but the rigorous EEG groups (Orozco Perez, Ingendoh) have been increasingly skeptical.
Clinical: Anaesthesia and perioperative nursing have shown the most clinical interest because the use case (10-minute pre-op anxiety reduction) is well-matched to the modality and effect size. No major guideline body (AASM, APA, NICE) endorses binaural beats for sleep, anxiety, or depression.
Community: Large self-experimenter community on Reddit, YouTube, and meditation forums. Thematic reports consistent in direction (relaxation, focus). Strong survivorship and confirmation bias — users who notice effects post and recommend; users who notice nothing simply stop using and don't post.
Skeptic / counter: The systematic-review wave (Ingendoh 2023, Basu & Banerjee 2023, Platt & Hammond 2024) has converged on "promising but methodologically weak, central entrainment claim unsupported, effect sizes shrink under blinding". No active anti-binaural-beats lobby, but the rigorous EEG community has effectively deflated the marketing claims through publication.

Population variability

Beat perceptibility itself varies: Oster noted that some listeners cannot perceive binaural beats at all, and that perceptibility varies by gender, by hormonal phase in women, and is reduced in some neurological conditions including Parkinson's disease (Oster 1973). Effect sizes in trials show substantial individual variance — meta-analytic g = 0.45 means many participants saw little effect even when groups did (Garcia-Argibay et al. 2019). Anxiety-reduction effects appear most robust in populations with elevated baseline anxiety (perioperative, chronic anxiety) and weaker in healthy non-stressed populations. Sleep effects in subclinical-insomnia adults were minimal under blinding (Bang et al. 2019); whether they would be larger in clinically diagnosed insomnia is untested. Age, hearing-acuity status (high-frequency hearing loss could affect carrier perception in older adults), and prior expectation (someone primed to expect benefit shows more benefit) all matter. No clear gender-stratified effect for outcomes, but the trial populations are small enough that subgroup analyses are underpowered.

Knowledge gaps

Double-blind sham-controlled trials at scale. The field is dominated by small (n < 50) single-site trials with inadequate blinding. A multi-site RCT (n > 300) with rigorous sham control (e.g., audio identical except for beat amplitude, blinded outcome assessors) for anxiety or sleep outcomes would settle the effect-size question.
Cortical entrainment vs. behavioural outcome. The literature has not adequately linked the (modest) cortical entrainment that does occur to behavioural and subjective outcomes. Studies that simultaneously measure EEG entrainment and behavioural change in the same participants would clarify whether the mechanism marketing relies on is doing the work.
Adverse-event reporting. CONSORT Harms 2022-compliant reporting is essentially absent from this literature (Platt & Hammond 2024). Wahbeh's signal of increased depression and poorer recall in some participants warrants systematic follow-up.
Optimal parameters. Carrier-tone frequency, beat-amplitude depth, masking-noise presence/absence, duration, pre- vs. during-task timing — meta-analyses suggest these matter but the parameter space is under-explored.
Long-term effects. Almost no trials beyond 60 days. Whether daily multi-month use produces tolerance, habituation, or cumulative benefit is unstudied.
Sleep architecture. The 2024 0.25 Hz sleep-onset trial and Dabiri's 3 Hz pilot are small and unblinded. Polysomnography-based, sham-controlled trials in clinical insomnia populations are missing.

Scope vs. brief. The brief named focus, relaxation, sleep onset, anxiety, meditative states, and evidence quality. All five effects plus the evidence-quality lens are addressed in the article. Anxiety reduction gets the most space because it carries the only replicated clinical signal; meditative-state effects get the least because the trial literature on meditative depth specifically is essentially absent — that's noted in the evidence section rather than hidden.

Action choice. Settled on know with cadence as-needed rather than do. The dominant reader takeaway is calibration — understanding what binaural beats actually do vs. what the consumer industry sells — and the (modest) protocol is offered as a footnote to that calibration. do would have overstated the case; avoid would have understated it (there is a real, if small, effect, and the intervention is essentially free and harmless).

Rating difficulties. Mood at 2 was the closest call. Anxiety reduction is the substance's strongest effect and the perioperative trials are reasonably clean, but effect sizes shrink under blinding and most of the lift is plausibly attributable to general calming audio plus expectancy. Settled on 2 (real but small contribution) rather than 3 (meaningful named effect) because the music-vs-beats delta in trials is usually narrow. Evidence at 2 reflects the contested EEG picture; controversy at 3 reflects the gap between the rigorous EEG community (skeptical) and the clinical anaesthesia + consumer communities (positive).

Contraindications. Considered flagging epilepsy. Deliberately did not, because (a) the closed-vocabulary token list doesn't include an epilepsy entry, (b) the empirical signal for auditory entrainment provoking seizures is much weaker than for photic stimulation, and (c) Platt & Hammond 2024 explicitly note that adverse-event reporting in this literature is essentially absent — there is no actual signal to anchor a contraindication to. The honest framing is "low-risk in non-clinical populations, theoretical concern for musicogenic / reflex epilepsy, no contraindication in the catalogue's closed vocabulary applies".

Separate-entry candidates. Isochronic tones / monaural beats could warrant their own entry if the catalogue grows in this direction — they are a meaningfully different stimulus class with their own (smaller) literature. Transcranial alternating-current stimulation (tACS) is a separate-entry candidate but probably belongs under medical/research rather than mental.

Future links. Once entries exist, link to: meditation / mindfulness practice, breathwork (4-7-8, box breathing), CBT-I for insomnia, ambient-audio / soundscape entries, and the Light & Environment entry on light exposure for circadian alignment (alternative for sleep onset).

Cite usage vs. dossier. Article uses 9 of the 11 citations added. Dossier-only refs: Wahbeh et al. 2007 (used in dossier for mechanism-failure-to-find-entrainment and the adverse-direction depression signal, less load-bearing in the article) and Baseanu et al. 2024 (used in evidence section but lighter weight). Dossier-as-superset property holds.