Perception
Hearing Range Test
Tones rise step by step from 8 kHz; you say whether you heard each one. Your result is the highest tone you reported hearing — which is a fact about your headphones and your volume as much as about your ears, and this page treats it that way.
Hearing range
You will hear a series of tones, each higher than the last, starting at 8 kHz. After each one, say whether you heard it. The test stops at the first tone you cannot hear, and your result is the highest one you could.
- Use headphones. Laptop and phone speakers fall away sharply above roughly 10 kHz, so on speakers this measures the speaker, not you.
- Set your volume now and do not change it. Turning it up mid-run raises the frequency you reach without telling us, and the result becomes meaningless.
- Some tones will be silent. They are there on purpose. Answer honestly — the results panel tells you how you did on them.
What your equipment contributed
- Audio output
- not measuredmeasured when you start
- Highest tone we will play
- —derived from the sample rate
- Headphones and volume
- unmeasurableand they bound the result further
Digital audio cannot represent a frequency at or above half the sample rate, so your device's output rate sets a hard ceiling on what we can play at all — we derive it from the rate your browser reports rather than assuming one, and we stop short of it because a tone right at that boundary is reconstructed differently by every DAC. The larger limit is the one we cannot measure: your headphones, your volume, and your room. Most consumer transducers roll off well below 20 kHz, and a tone played 10 dB quieter is a tone you stop hearing sooner. So the honest reading of your result is a floor — you hear at least this high, on this equipment, at this volume. It is not the highest frequency your ears can detect, and no browser test can give you that number.
How this is measured
What starts and stops the run
The ladder runs 8, 10, 12, 14, then 1 kHz steps to 20 kHz — coarse through the range essentially everyone hears, fine where adult limits actually land. Each tone plays for 1.5 seconds and you answer heard or nothing. The run ends at your first “nothing”, and your result is the highest step before it. Because the steps near the top are 1 kHz apart, your true limit sits somewhere between the last tone you heard and the first you didn't — we cannot place it closer than that, and we say so on the readout rather than reporting a precision we don't have.
Why the tones fade in and out
This is the detail that decides whether the test measures anything. Switching a tone on abruptly is a step change, and a step contains energy at every frequency — so an abruptly gated 19 kHz tone produces a broadband click that is plainly audible to someone who cannot hear 19 kHz at all. A test built that way measures how well you hear clicks and reports it as your hearing range. Every tone here ramps up and down over 60 ms, which is far too gradual to click, so the only thing there is to hear is the tone itself.
The silent steps, and why they are there
Two steps in every run play nothing. They are placed low in the ladder, where nearly everyone is still comfortably hearing tones, because the run stops at your first “nothing” — a check placed near the top would never be reached and would quietly do nothing. Reporting that you heard a silent step doesn't end the run, but it does mean we cannot tell which of your other answers were tones and which were expectation, so the results panel says the number is unreliable and why. Discarding that silently, and showing you a confident figure anyway, is exactly the thing this site exists not to do.
What we deliberately don't show you
The frequency of the tone currently playing. Knowing that you are “on 17 kHz now” changes what you think you heard, and this test is entirely dependent on self-report. You see the full ladder — every step, every answer, and both silent steps — on the results panel once the run is over.
What we cannot tell
Whether you changed the volume mid-run, what your headphones do above 15 kHz, how quiet your room is, or whether the faint thing you reported at the top of the ladder was the tone or your own expectation. Nothing is timed here, so there is no reaction-time component and no display involvement — the result is a frequency, not a duration. Nothing you do on this page is stored or transmitted.
How your result compares
Extended high-frequency audiometry is well studied, so a defensible reference range exists — with a large caveat we will get to. In a sample of 162 otologically healthy adults aged 21–70, tested with calibrated audiometers and high-frequency earphones, the share of ears responding at all fell steeply with both frequency and age:
| Age | 16 kHz | 18 kHz | 20 kHz |
|---|---|---|---|
| 21–30 | 100% | 78% | 52% |
| 31–40 | 64% | 20% | 8% |
| 41–50 | 39% | 2% | 2% |
| 51–60 | 31% | 3% | 0% |
| 61–70 | 16% | 0% | 0% |
The pattern is the useful part: hearing above 16 kHz is uncommon past the thirties, and no one in the two oldest groups responded at 20 kHz. If you are 45 and stopped at 14 kHz, you are not unusual.
What we matched, and what we could not
Matched: the frequencies, and the ascending order.
Not matched — and this one is large: that study measured a threshold. It raised the level at each frequency until the listener responded, and reported how loud that had to be. Median thresholds at 16 kHz ran from about 25 dB HL in the youngest group to around 55 dB HL in the middle groups — meaning many of those responders needed a loud tone. We play one fixed level and ask yes or no, so a listener who would have responded at 55 dB HL may simply not hear our tone. Their equipment was calibrated to a standard; ours is whatever you have on your head. Their room was treated; yours is not.
So the table is context for the shape of the decline, not a line you have passed or failed. Our result is a floor — you hear at least this high, on this equipment, at this volume — and a floor cannot be compared like-for-like against a threshold.
No percentile is shown, and on this test we would need two separate things we don't have to justify one: the distribution of the published measurements, and a way to relate our fixed-level yes/no to their threshold search. Neither exists. A percentile drawn from BlinkBench's own visitors would be worse still — it would be a ranking of headphones as much as of hearing, computed from a self-selected sample we cannot verify.
Source: Wang, M., Ai, Y., Han, Y., Fan, Z., Shi, P., & Wang, H. (2021). Extended high-frequency audiometry in healthy adults with different age groups. Journal of Otolaryngology — Head & Neck Surgery, 50, 52. doi:10.1186/s40463-021-00534-w
Last reviewed: July 2026
Frequently asked questions
Why did some of the tones play nothing at all?
Because the only thing this page can observe is you clicking a button, and near your limit the difference between a faint tone and expecting a faint tone is genuinely hard to call. Two of the steps in every run are silent. If you report hearing one, we tell you on the results panel and we tell you the number is unreliable — we don't quietly drop it and show you a score anyway. Almost every other version of this test online has no such check, which means it has no way of knowing whether you heard anything at all.
Why is my result lower than the hearing-age charts say it should be?
Most likely your headphones or your volume, not your ears. Consumer transducers roll off well before 20 kHz, and the level a tone is played at strongly affects whether you detect it — proper audiometry finds the quietest level you respond to at each frequency, whereas we play one fixed level and ask yes or no. The charts that convert a frequency into a hearing age are also doing something we won't: treating an uncalibrated browser measurement as if it were a clinical one. Your result is a floor. It says you hear at least this high on this equipment.
Can this test detect hearing loss?
No. It measures one thing — the highest frequency you reported hearing at one uncalibrated level — and hearing loss is mostly diagnosed in the speech frequencies well below where this test even starts. You can have significant hearing loss and pass every step here, and you can hear normally and stop early because your earbuds do. It is not a screening tool and cannot be used as one. If you have any concern about your hearing, see an audiologist, who will measure the quietest level you respond to at each frequency using calibrated equipment in a treated room.
Why does the test stop at 20 kHz, or lower on my device?
Because digital audio cannot represent a frequency at or above half the sample rate, and we won't play a tone we can't vouch for. Your browser reports its output sample rate; we derive the ceiling from that number rather than assuming one, and we stop short of it because a tone right at that boundary is reconstructed differently by every DAC. On a 48 kHz device the ladder reaches 20 kHz; on a 44.1 kHz device it stops at 19 kHz. If you reach the top step, the honest reading is that you reached the top of our ladder — not the top of your hearing.
Related tools
The smallest colour difference you can reliably find, reported as three separate thresholds in ΔE — lightness, green–red and blue–yellow — and never finer than your display can render.
Open tool →Reaction Time TestFive trials measuring how fast you respond to a colour change, reported as a median with the share your own display could account for.
Open tool →Number Memory TestThe longest string of digits you can hold and recall, presented one per second to match the research the score is compared against.
Open tool →