Hearing Range by Age
Play the same test tone to a teenager and their parent in the same room, and it's quite possible one of them winces while the other checks whether the speaker is broken. Neither is wrong, and neither device is at fault. The top of our hearing range slips down with age for almost everyone, so slowly that nobody notices it happening to their own ears. That's why the gap between a fifteen-year-old and a fifty-year-old can be a whole octave.
You lose the top of the range, not the middle
This gradual, age-related change has a name: presbycusis. Wikipedia's summary, drawing on the medical research, describes it as affecting high pitches before low ones, and says measurable change can start as early as 18, long before anyone would say they have "hearing loss". The World Health Organization's fact sheet on deafness and hearing loss shows where the same process ends up: over 25% of people older than 60 are affected by disabling hearing loss. Those two facts describe the same slope from opposite ends. One says it starts early; the other says how far it has gone by the time it affects a quarter of an age group.
What happens to the top of the range, decade by decade
Studies that test pitches above the usual hearing-test range (known as extended high-frequency testing) show the decline directly. A 2021 study of 162 adults with healthy ears at Shandong Provincial ENT Hospital in China measured the median threshold at 16 kHz for each decade of adult life, and it's a steep curve rather than a straight line:
| Age | Median threshold at 16 kHz | What that means in practice |
|---|---|---|
| 21-30 | 25 dB HL | Every subject under 30 in the study could respond to the tone at a level close to a young, healthy ear. |
| 31-40 | 50 dB HL | The tone has to be about twenty-five decibels louder to be heard. That's the biggest jump between any two neighbouring decades in the study. |
| 41-50 | 55 dB HL | A small further rise on top of the thirties. |
| 51-60 | 55 dB HL | The median stays about the same, but see the note below: fewer people in this group could hear the tone at all. |
| 61-70 | Response rate at 16 kHz fell to 16.1%, too few people for a median to mean much. | |
A separate study, published in Audiology Research and available on PubMed Central, split adults more broadly, 18-40 versus 40-70, but tested more pitches across the top of the range. It found the two groups close together at 9 kHz (5 dB HL versus 20 dB HL median) and far apart by 14 kHz (17.5 dB HL versus 70 dB HL median). The two studies divide age up differently but agree on the shape: the decline is small and slow lower down and steep at the very top, which is exactly the octave or two behind the "can you still hear that?" party trick.
Why 20 kHz is a young person's ceiling, not everyone's
The "20 Hz to 20 kHz" printed on speaker boxes and headphone specs describes the outer edge of what a healthy young ear can hear in good conditions. It was never a promise to every listener, at every age, for life. Even that is the extreme case. A 2022 survey of adults aged 18-29 with no hearing complaints found a median threshold at 18 kHz of about 17 dB HL in both ears, already well above the near-zero thresholds the same people had at 8-9 kHz. In other words, the slope toward the ceiling has already started in ears young enough to be the textbook example. The last kilohertz or two before 20 kHz has always been the thinnest part of the range: there when you're young, and gone early for most people.
The best-known demonstration is the "mosquito tone", a sound at about 17.4 kHz used in, among other things, a commercial device designed to be heard mainly by young people and sold on that basis. Wikipedia's account of the device, summarising the manufacturer's own specifications rather than an independent hearing study, says the tone is typically heard only by people under about 25. That's a manufacturer's claim passed on by an encyclopedia, not a medical measurement. It's a good illustration of how quickly the top end narrows, but not a cut-off for any one person.
A range, not a cut-off
Every figure above is a median or a response rate across a group, not a wall you hit on a particular birthday. The 51-60 group in the Shandong study had a similar median to the 41-50 group, but fewer of them could hear the tone at all. The group was splitting between people still close to the young norm and people who couldn't hear 16 kHz anymore. A single number for "what people your age can hear" hides exactly the spread that decides which of those you are. The formal way to describe that spread, the full statistical range of hearing against age rather than just a median, is ISO 7029, the international standard on the subject. It's behind ISO's paywall, so the figures here come from the published studies above rather than the standard itself.
A median isn't a diagnosis, and this page isn't trying to give one. Hearing loss also has causes that have nothing to do with age, loud noise being the biggest, and a test run in a browser through headphones nobody has measured can't tell "getting older" from "too many loud gigs" from "these earbuds just drop off early". If a result worries you, a specialist has calibrated equipment and a sound booth; this page and its tool don't. What our hearing range test can give you is one honest, repeatable result from your own ears and your own equipment. Not a curve and not a verdict, but a real number to compare with your own past results, rather than with a stranger's decade.
Age is one part of how high a sound has to be before you stop hearing it; loudness is the other, and it has its own risks at any pitch. How Loud Is Too Loud covers what actually damages hearing at any age. If what brought you here is really a Bluetooth pair sounding thin at the top, Bluetooth Audio Latency by Codec covers a different way high notes get lost before they reach your ears. And if a speaker has gone silent rather than just thin, that's a hardware question this page can't answer. Dead, Blown or Miswired: Diagnosing a Silent Speaker explains how to tell a dead speaker from a damaged one or a wiring fault.
Every figure above links to the study it came from. Where a claim is a manufacturer's specification passed on by another source, the paragraph says so, rather than crediting a study or standard it didn't come from.