Audiometric testing is a family of clinical measurements used to determine how well you hear across different pitches and volumes, pinpoint where in the auditory system a problem lies, and track changes over time. The most familiar version is pure-tone audiometry, where you sit in a quiet booth wearing headphones and press a button whenever you hear a beep. But a full audiometric evaluation can involve half a dozen distinct tests, from speech recognition tasks to measurements of middle-ear pressure to recordings of electrical activity along the auditory nerve. What ties them all together is a shared goal: mapping your hearing with enough precision that a clinician can figure out not just whether you have hearing loss, but what kind, how severe, and what to do about it.
Pure-Tone Audiometry and the Hughson-Westlake Method
The backbone of any hearing evaluation is the pure-tone test. An audiometer generates tones at specific frequencies, typically from 250 Hz (a low hum) up through 8,000 Hz (a high-pitched whistle), and presents them through headphones at varying loudness levels measured in decibels. Your job is simple: respond every time you hear a tone, no matter how faint. The audiologist uses a bracketing procedure called the modified Hughson-Westlake method, which raises and lowers the volume in small steps, usually 5 dB at a time, to find the quietest level at which you can detect each pitch at least half the time.1Ear and Hearing. High-Frequency Pure-Tone Audiometry in Children: A Test–Retest Reliability Study Relative to Ototoxic Criteria That quietest level is your hearing threshold for that frequency.
The result is plotted on an audiogram, a graph with frequency along the top and loudness down the side. Normal hearing thresholds fall at or near 0 to 25 dB across the frequency range. If your thresholds are higher (meaning you need a louder sound to hear it), that indicates hearing loss at those pitches. The shape of the audiogram tells the clinician a lot: a dip at high frequencies is the classic signature of noise damage, while a flat pattern across all pitches suggests a different cause. Yet despite how straightforward audiograms look, expert audiologists frequently disagree on how to categorize the configuration and severity of any given audiogram, which is why the raw numbers matter more than the label attached to them.2PubMed Central. Toward a standard description of hearing loss
What Happens Before You Put on the Headphones
Before any tones play, a clinician will typically look into your ears with an otoscope. This is not just a formality. Impacted earwax is surprisingly common and can meaningfully affect results. In a study of hospitalized older adults, about 35% had impacted cerumen, and after it was removed, hearing improved in 75% of the affected ears, with some people able to hear several additional tones they had previously missed.3PubMed. Effect of cerumen removal on the hearing ability of geriatric patients If wax is blocking the ear canal, anything measured after that point would overestimate the degree of hearing loss.
The testing environment matters, too. Clinical audiometry is performed inside a sound-treated booth designed to block external noise. The American National Standards Institute specifies maximum permissible ambient noise levels for audiometric rooms, and these limits are strict enough that testing down to 0 dB hearing level is possible only if background noise is controlled with appropriate booth construction and headphone attenuation.4PubMed. Ambient noise levels in audiometric test rooms used for clinical audiometry Even with a proper booth, certain headphone cushions alone may not provide enough isolation to meet these standards, which is why noise-excluding enclosures or insert earphones are sometimes used.5PubMed. Real ear attenuation at threshold for three audiometric headphone devices: implications for maximum permissible ambient noise level standards The audiometer itself must also be calibrated against reference sound pressure levels so that 0 dB on the dial corresponds to a known physical intensity, ensuring that results are comparable from one clinic to another.6PubMed Central. Reference equivalent threshold sound pressure levels for the Wireless Automated Hearing Test System
Bone Conduction and Telling the Types of Hearing Loss Apart
Standard headphones deliver sound through the air into the ear canal, which is called air conduction. But there is a second route: a vibrating device placed on the bone behind your ear sends sound directly through the skull to the inner ear, bypassing the eardrum and middle ear entirely. This is bone conduction testing, and comparing the two sets of thresholds is how clinicians distinguish between the main types of hearing loss.
If both air and bone thresholds are elevated by roughly the same amount, the inner ear (or the nerve pathways beyond it) is the problem. That is sensorineural hearing loss. If air-conduction thresholds are worse but bone-conduction thresholds are normal or near-normal, the bottleneck is somewhere in the outer or middle ear, and you have conductive hearing loss. A gap between the two, called the air-bone gap, tells the clinician exactly how much of the loss is mechanical versus sensory. Some people have both, which is called mixed hearing loss. Research has shown that combining pure-tone data with speech-in-noise testing can distinguish conductive from sensorineural loss with very high accuracy, but the traditional air-bone comparison remains the standard clinical approach.7Taylor & Francis Online / PubMed Central. Pure-tone audiometry without bone-conduction thresholds: using the digits-in-noise test to detect conductive hearing loss
The Masking Problem
Here is something most people never think about: when you test one ear with a loud tone, the sound can travel through the skull and be picked up by the other ear. This is called cross-hearing, and it becomes a real issue when there is a large difference in sensitivity between the two ears. If the better ear “hears” the tone meant for the worse ear, the audiogram for the worse ear looks artificially good.
To prevent this, audiologists deliver a controlled noise signal, called masking noise, to the better ear during testing. The noise keeps the better ear busy so it cannot eavesdrop on tones being sent to the worse ear.8PubMed Central. Dynamically Masked Audiograms With Machine Learning Audiometry Getting the masking level right is one of the trickier parts of clinical audiometry: too little noise and the better ear still picks up the test tone; too much and the masking noise itself crosses over and interferes. This is part of why a trained audiologist is not just pressing buttons in sequence. The decision tree around when to mask, and how much, requires real-time clinical judgment.
Speech Audiometry
Hearing pure tones is one thing. Understanding speech is another, and the gap between the two is often what brings people into the clinic in the first place. Speech audiometry tests how well you can recognize and repeat spoken words at different volume levels.
There are two main measurements. The speech reception threshold finds the quietest level at which you can correctly repeat simple two-syllable words about half the time. It should roughly agree with your pure-tone thresholds; a big mismatch is a red flag that something unusual is going on. The word recognition score tests how accurately you can identify single-syllable words presented at a comfortable listening level. This score is expressed as a percentage and gives a sense of how much benefit you might get from hearing aids or other amplification.9PubMed. The development of the University of Jordan word recognition test Someone with a high word recognition score and mild loss is a great candidate for hearing aids. Someone with poor word recognition even at loud levels may have nerve damage that amplification alone cannot fix.
Middle Ear Tests
Pure-tone and speech audiometry tell you what the ear can and cannot hear, but they do not directly measure what is happening in the middle ear. Tympanometry fills that gap. A small probe inserted into the ear canal changes the air pressure while emitting a low-frequency tone and measuring how much sound energy bounces back off the eardrum. A healthy eardrum reflects the least energy when the pressure on both sides is equal; a graph of that response (the tympanogram) peaks around normal atmospheric pressure. A flat tympanogram can indicate fluid behind the eardrum, and a shifted peak might suggest eustachian tube problems. Research on middle-ear mechanics has confirmed that the compliance of the tympanic membrane (how easily it moves) correlates with static admittance measurements, and that the ear canal pressure reading at the eardrum’s resting position tracks middle-ear pressure.10Elsevier / PubMed Central. Mechanics of the middle ear system: computerized measurements of its pressure-volume relationship
Related to tympanometry is acoustic reflex testing. Loud sounds normally trigger a tiny muscle in the middle ear, the stapedius, to contract and stiffen the chain of bones that conduct sound. By measuring the change in ear canal impedance during a loud tone, the audiologist can determine whether this reflex is present, absent, or abnormal. The reflex threshold and the way the reflex decays over time are diagnostically useful. In patients with tumors on the hearing nerve, for instance, reflex decay measured at higher sensation levels has been shown to be more sensitive and specific for detecting the tumor than decay measured at the conventionally used level.11PubMed. Stapedius reflex quantification in acoustic tumor patients The reflex arc also passes through the brainstem, so abnormalities can point to problems higher up the pathway than the ear itself.
Objective Tests That Do Not Require a Button Press
Not every patient can raise a hand or press a button when they hear a sound. Newborns, people who are unconscious, and some individuals with cognitive disabilities need objective measures. Two tests dominate this space.
Otoacoustic emissions (OAEs) are faint sounds generated by the outer hair cells of the inner ear in response to incoming sound. A tiny microphone in the ear canal picks them up. If the outer hair cells are healthy, emissions are present; if they are damaged, emissions are reduced or absent. Because these cells are often the first casualties of noise exposure and aging, OAE testing can catch damage even before the standard audiogram shows a problem. Research using distortion product OAEs has confirmed that outer hair cell loss is a contributing cause of age-related hearing loss regardless of the audiogram’s shape.12PubMed Central. A DPOAE assessment of outer hair cell integrity in ears with age-related hearing loss
The auditory brainstem response (ABR) takes things further up the nerve pathway. Electrodes on the scalp record the tiny electrical signals generated as sound travels from the cochlea through the auditory nerve and brainstem. The recording shows a series of waves, numbered I through V, each corresponding to a specific station along the auditory pathway. Clinicians analyze the latency, amplitude, and spacing of these waves to evaluate the integrity of the entire route from the inner ear to the midbrain.13PubMed. Auditory brainstem response: Key parameters for good-quality recording ABR is the gold standard for newborn hearing screening and is also used to estimate hearing thresholds in anyone who cannot participate in behavioral testing.
Testing Children
Audiometry with kids requires creative adaptation, because asking a six-month-old to press a button when they hear a beep is not going to work. Clinicians use age-appropriate behavioral techniques that turn the test into something engaging enough to hold a child’s attention.
For infants roughly six months to about two years old, the standard approach is visual reinforcement audiometry (VRA). The child sits on a parent’s lap while sounds play from a speaker or insert earphone. When the child turns toward the sound, a visual reward appears: an animated toy lights up, or a video clip plays on a screen. The child quickly learns the association between sound and reward, and the audiologist can map out thresholds by observing the head-turn response. Research into optimizing VRA has found that using a greater variety and complexity of visual rewards, keeping reward durations short, and offering breaks all help sustain the child’s attention. Interestingly, younger infants around 12 months tend to produce more usable responses than toddlers around 18 to 24 months, possibly because older children habituate faster and lose interest more quickly.14PubMed. Optimisation of visual reinforcement audiometry: a scoping review Specialized reinforcer libraries have even been developed for children with developmental disabilities, featuring themed photo and video sets based on interests commonly reported for that population.15PubMed Central. Visual Reinforcers Designed for Children with Developmental Disabilities
Once children are around two and a half to five years old, conditioned play audiometry (CPA) takes over. Here the child is taught a play task, like stacking a ring on a peg every time they hear a tone. After a few demonstrations, the child can be conditioned to respond reliably, and the audiologist can obtain frequency-specific and ear-specific thresholds through both air and bone conduction.16Scientific Reports. Behavior recognition technology based on deep learning used in pediatric behavioral audiometry By around age five or six, most children can cooperate with conventional hand-raise or button-press audiometry.
Extended High-Frequency Audiometry
Standard audiometry covers frequencies up to 8,000 Hz, but the human ear can detect sounds up to about 20,000 Hz (though this ceiling drops with age). Extended high-frequency audiometry tests thresholds at frequencies from 9,000 to 20,000 Hz and can reveal damage that conventional audiometry misses entirely. This matters because the highest-frequency hair cells in the inner ear tend to be the first to go when exposed to noise or ototoxic drugs like cisplatin-based chemotherapy. A person can have a completely normal standard audiogram while already showing significant threshold shifts at these extended frequencies.17Acta Otorrinolaringologica (English Edition). Extended High-frequency Audiometry (9000–20 000 Hz). Usefulness in Audiological Diagnosis
For workers exposed to hazardous noise, extended high-frequency testing during routine surveillance exams can flag damage years before it shows up on a standard audiogram, potentially allowing earlier intervention.18Noise and Health. High-frequency audiometry: A means for early diagnosis of noise-induced hearing loss Despite its value, extended high-frequency testing is not yet a routine part of most clinical evaluations, partly because normative data is less well established and test-retest variability increases at very high frequencies.
Workplace Hearing Conservation Programs
If you work around loud equipment, you probably get your hearing tested periodically as part of an occupational hearing conservation program. These programs are regulated by OSHA and focus on detecting shifts in hearing before they become disabling. The key metric is the standard threshold shift (STS), defined as an average change of 10 dB or more at 2,000, 3,000, and 4,000 Hz in either ear compared to your baseline audiogram.
A recordable hearing loss under OSHA rules typically shows up after hearing has already moved into impairment territory, while the non-age-corrected STS criterion usually flags a change before that point.19PubMed Central. Impact of OSHA final rule–recording hearing loss: an analysis of an industrial audiometric dataset In other words, the STS is designed as an early warning system, not a measure of existing disability. There has been debate about whether the OSHA criterion is sensitive enough, with NIOSH recommending that alternative threshold-shift criteria replace or supplement the STS for workplace surveillance.20Journal of Occupational and Environmental Medicine. Comparison of Criteria for Significant Threshold Shift in Workplace Hearing Conservation Programs
The scale of the problem is large. In a study spanning 2001 to 2010, about 20% of workers showed a NIOSH-defined significant threshold shift. Men were about 1.4 times more likely than women to experience one, and workers aged 56 to 65 were nearly four times more likely than those aged 18 to 25.21PubMed Central. Prevalence of Workers with Shifts in Hearing by Industry: A Comparison of OSHA and NIOSH Hearing Shift Criteria
Smartphone Audiometry and Remote Testing
The pandemic accelerated interest in moving hearing tests out of the booth and onto personal devices. Smartphone-based audiometry apps have proliferated, and the question is whether they are accurate enough to trust. A meta-analysis of studies using smartphone apps found sensitivity around 89% and specificity around 93% for detecting hearing loss with pure-tone-based approaches, and comparable performance for speech-based methods.22PubMed Central. Diagnostic Accuracy of Smartphone-Based Audiometry for Hearing Loss Detection: Meta-analysis That is respectable as a screening tool but not quite the precision needed for clinical diagnosis.
When home-based app results were compared to clinical audiometry head to head, thresholds did not differ significantly at individual test frequencies, and test-retest reliability was comparable to what you get in a sound booth.23PubMed. Home-Based Audiometry With a Smartphone App: Reliable Results? The catch is that accuracy drops for people who actually have hearing loss, especially asymmetric loss (one ear worse than the other). One study found that while the app correctly classified normal hearing about 97% of the time, it caught only about 22% of mild hearing losses and about 39% of more significant losses, and results for the worse ear in asymmetric cases were particularly unreliable.24Scientific Reports. Assessing the accuracy and reliability of application-based audiometry for hearing evaluation That is a problem, because the people most likely to use a hearing screening app are precisely the people who suspect they have a loss.
The underlying issue is environmental noise. You can control what happens inside a sound booth; you cannot control a user’s living room. Uncalibrated consumer headphones add another layer of uncertainty. For now, smartphone apps are reasonable for flagging whether you should seek a professional evaluation, but they are not a replacement for one.
How Mental State Can Affect Your Results
Audiometric testing seems like a straightforward physical measurement, but it depends on your active participation and attention. Research has found that cognitive load, basically being mentally distracted, can push pure-tone thresholds up by about 2 dB. In younger adults, the effect appeared only when the distraction involved auditory processing (like mentally rehearsing a list of words), but older adults showed elevated thresholds under both auditory and visual distractions.25PubMed Central. Effects of Cognitive Load on Pure-Tone Audiometry Thresholds in Younger and Older Adults Two decibels is not a dramatic shift, but it is real, and it means test conditions that minimize distraction and anxiety genuinely improve accuracy. If you are asked to put your phone away and sit quietly, that is not just clinic protocol. It is protecting the quality of your data.
Central Auditory Processing Tests
Some people hear perfectly on a standard audiogram but struggle to understand speech in noisy environments, follow rapid conversation, or distinguish between similar-sounding words. When the peripheral hearing organs check out fine but comprehension problems persist, clinicians may investigate central auditory processing, the brain’s ability to make sense of what the ears deliver. Testing for this involves a separate battery of tasks that stress the auditory system in specific ways: listening to speech filtered to remove high frequencies, hearing different words in each ear simultaneously (dichotic listening), or identifying patterns of tones that differ in pitch or duration.26PubMed. Normative values for tests of central auditory processing disorder in children aged from 6 to 12 years old
Diagnosing central auditory processing disorder is more contentious than diagnosing peripheral hearing loss. The criteria used to define abnormal performance vary between clinics and researchers, and studies have shown that the number of children diagnosed changes dramatically depending on which scoring criteria are applied to the same test results.27PubMed. Using different criteria to diagnose (central) auditory processing disorder: how big a difference does it make? This does not mean the condition is not real. It means the field has not yet settled on where exactly to draw the lines, and results need to be interpreted with some humility about the state of the science.