How Old Are Your Ears?

Treat it as entertainment with a grain of truth, not a measurement. Two things limit it badly. First, the age mapping it uses — like every "how old are your ears" test online — isn't published in any standard or study; the version everyone repeats comes from a 2013 YouTube video with no source. Second, every tone plays at the same volume, but your ear needs roughly 38 dB more sound pressure to detect 16,000 Hz than 8,000 Hz, so the point where tones stop being audible is partly an artifact of that fixed volume. Add Bluetooth codecs that cut off between about 14,200 and 18,900 Hz depending on your phone, and a "no" at the top may be your headphones rather than your ears. It's a fun comparison, not a diagnostic.

High-frequency loss is a normal part of ageing, but the usual "those hair cells are more delicate" explanation isn't quite the mechanism. The cochlea is arranged tonotopically — high frequencies are detected at its base, near the entrance, and low frequencies deep at the apex. The base takes the brunt of a lifetime of sound energy because everything entering the ear passes through it, so cumulative exposure, reduced blood supply, and loss of outer hair cells hit there first. Noise exposure, genetics, ototoxic medications, and cardiovascular health all accelerate it. There's also a documented sex difference: women tend to retain better extended high-frequency hearing than men, by roughly 9 dB in older adults — one more reason a single tone can't tell you an age.

Once high-frequency hearing cells are damaged, they don't grow back. But you can absolutely protect what you have. Wearing hearing protection in loud environments, keeping earphone volumes moderate, and staying on top of your overall health all help preserve your hearing for years to come.

Earlier than most people expect, but more gradually than the internet suggests. What fades in your twenties is the band above 16,000 Hz — on calibrated equipment only about 78% of 21-to-30-year-olds respond at 18,000 Hz, and only half at 20,000 Hz. At 16,000 Hz itself, essentially every adult under 30 still hears it, with the first clear group-level decline showing up in the 31–40 bracket. Loss reaching the conversation frequencies is a different and much later story: around 5% of adults aged 45–54 and 10% of those 55–64 have disabling hearing loss, so for most people in their forties and fifties, speech-range hearing is still within normal limits.

Enormously — and the culprit is usually the wireless connection rather than the earbuds themselves. Measurements of Bluetooth audio show AAC rolling off at about 14,200 Hz on one phone, 16,000 on another, and 18,900 on an iPhone, while SBC at typical bitrates periodically cuts everything above 16,000 Hz outright. That means the same headphones can produce different "hearing ages" on different phones. Wired headphones remove that variable entirely, which is why we recommend them. But no browser test can calibrate an unknown pair of headphones — validated smartphone hearing tests use calibrated equipment, and uncalibrated setups have been shown to differ from proper audiometry by more than 14 dB on average. Quality helps; it is not the same as calibration.

Genuinely, no. The mapping this test uses isn't from any study, your headphones and phone shape the result as much as your ears do, and individual variation at these frequencies is enormous — plenty of people in their fifties still hear 16,000 Hz while some twenty-somethings don't reach 18,000 Hz. Even the international standard for age-related hearing norms says outright that you can't determine, for one individual, how much of a hearing change is down to age versus noise or other causes. A far better signal is your daily life: if you're missing conversations, asking people to repeat themselves, or struggling in restaurants, that's worth an audiologist appointment regardless of what this page said.

No — they measure different things. Apple's Hearing Test on AirPods Pro 2 and AirPods Pro 3 (paired with an iPhone on iOS 18 or later) plays tones from 250 to 8,000 Hz and produces a full audiogram of how quietly you can hear across the speech range. This hearing age test checks the opposite end: the very high frequencies from 8,000 to 18,000 Hz that fade first with age. Think of ours as a quick, fun screening and Apple's as a more clinical-style check — and neither replaces a visit to an audiologist.

A full audiogram measures the quietest sound you can hear at each frequency across the speech range, usually 250 to 8,000 Hz, and plots those thresholds in decibels. That's a measurement. This test only finds the highest frequency you can detect at one fixed volume, which is a fundamentally weaker thing — it produces a single cutoff rather than a curve, and that cutoff moves depending on your hardware. For a picture of how you actually hear conversation, start with our free hearing test and our audiogram guide.

Sometimes, and it's called extended high-frequency (EHF) audiometry — covering roughly 9,000 to 20,000 Hz. It isn't part of a routine audiogram because those frequencies contribute little to understanding speech, so a standard test stops at 8,000 Hz. Where EHF genuinely earns its place is early warning: it's used to monitor people taking ototoxic medications like certain chemotherapy drugs, because damage often appears up there before it reaches the speech range, and to catch early noise damage. If you're on a medication with hearing risks, EHF monitoring is worth asking about — our medication checker lists common culprits.

Not from audiology. It traces to the Mosquito, a device patented in 2005 that emits a roughly 17,400 Hz tone to deter teenagers from loitering, on the premise that people under about 25 could hear it and older adults couldn't. Students promptly turned it around into the "Teen Buzz" ringtone to text in class, and the device won an Ig Nobel Prize in 2006. The specific frequency-to-age table used by this and every similar test online descends from a 2013 YouTube video that cited no source — and different versions of that table disagree with each other, which tells you there's no shared dataset behind any of them.