Standard Threshold Shift: OSHA's Hearing Loss Rule

An annual audiogram comes back showing a 12 dB shift at the key frequencies, and someone on the safety team logs it as recordable on the spot. That's often wrong. A standard threshold shift osha determination has two separate hurdles, and clearing the first one — the shift itself — tells you nothing about the second. Miss that, and you either bury a real case or record one that never needed to go on the log. This piece walks through both parts of the test, a worked audiogram, and the age-correction step that trips up more people than anything else in 1904.10.

What is a standard threshold shift under OSHA rules?

A standard threshold shift, or STS, is a specific, measurable change in an employee's hearing compared with their baseline audiogram. It's defined as an average shift of 10 decibels (dB) or more at 2000, 3000, and 4000 Hz, in one or both ears, relative to that employee's baseline hearing test. The definition comes from the hearing conservation standard, and 1904.10 borrows it directly.

Those three frequencies aren't arbitrary. They sit in the range where occupational noise exposure does the most damage, and where hearing loss first shows up before it spreads into the frequencies used for everyday speech. An audiologist averages the dB level at those three points on the current test, averages the same three points on the baseline test, and compares the two. A 10 dB or greater increase, averaged across those three, is an STS.

That's step one. It is not, by itself, enough to make a case recordable.

Why does a shift alone not make a case recordable?

Under 1904.10, a hearing loss case is recordable only when both of the following are true in the same ear:

1. The employee has a work-related STS, as defined above.
2. The employee's total hearing level is 25 dB or more above audiometric zero, averaged at 2000, 3000, and 4000 Hz.

The second part is the absolute threshold test. It asks a different question than the shift test does. The shift test asks "did this person's hearing get worse compared with where they started?" The absolute threshold test asks "is their hearing now bad enough, in absolute terms, to matter?" An employee can have a real, work-related 10 dB shift and still not meet the 25 dB absolute floor — because their baseline hearing was excellent to begin with. In that case, the rule says: not recordable.

This is the part of 1904.10 that catches people out. A shift of 15 or 20 dB looks dramatic on the page. If the resulting average is still under 25 dB total, the case doesn't go on the log. OSHA built the rule this way on purpose — it's trying to capture hearing loss that has become clinically meaningful, not just any measurable change from a noisy testing booth on a given day. For general background on how these binary recording tests work across all injury types, see our OSHA recordable criteria checklist.

How is the STS average calculated at 2000, 3000 and 4000 Hz?

Take the hearing level in decibels at each of the three frequencies on the current audiogram, add them, divide by three. Do the same for the baseline audiogram. Subtract the baseline average from the current average. If the result is 10 dB or more in either ear, that ear has an STS. The calculation is done per ear — a shift in the left ear doesn't get averaged against a stable right ear to dilute the result.

A worked audiogram example

Here's how the maths plays out for a real-looking case. The employee is a press operator with five years on the line, routine noise exposure, and hearing protection issued but not always worn consistently — a detail that matters for context, not for the calculation itself.

Worked audiogram comparison for the right ear: baseline versus annual test at 2000/3000/4000 Hz, showing the average shift, whether it crosses the 10 dB STS threshold, and whether the case also meets the 25 dB absolute threshold.

The average shift is exactly 10 dB — that's an STS, since 10 dB or more is the trigger. The current average total hearing level is 30 dB, which clears the 25 dB absolute threshold too. Both parts of the test are met, in the same ear, so this case is recordable under 1904.10, assuming a physician or audiologist confirms the shift is work-related.

Now change one number. If the baseline had been 30/30/30 instead of 15/20/25, and the annual test came back 30/35/40, the shift average would be (0+5+15)/3 = 6.67 dB — under 10 dB, so no STS at all, and the case stops there regardless of the absolute level.

Or flip it the other way: a baseline of 5/5/10 and an annual test of 15/15/20 gives a shift average of exactly 10 dB — a genuine STS — but the current average is only 16.67 dB, well under the 25 dB floor. Real shift, confirmed by the numbers, and still not recordable, because the employee's hearing simply started from a very good place. That's the case that generates the most pushback internally, because "the audiologist flagged an STS" sounds final to people who haven't read the second half of the rule.

Two contrasting examples from the same plant

Case A — recordable. A forklift operator's baseline, taken at hire, averaged 12 dB across the three frequencies. Three years later, the annual audiogram averages 24 dB — a 12 dB shift, comfortably over the STS threshold. The current average of 24 dB is just under 25, though, so on paper it narrowly misses the absolute threshold. But the audiologist applies age correction (more on that below), which in this case pushes the current-test contribution down slightly rather than up, and the adjusted average lands at 26 dB. Both tests are now met. Recordable.

Case B — not recordable, on review. A maintenance technician's audiogram shows a 14 dB shift and a current average of 28 dB — on first read, both thresholds look cleared. But the reviewing physician determines the technician had an ear infection with temporary fluid build-up at the time of testing, unrelated to noise exposure, and a retest six weeks later shows hearing back near baseline. The original test doesn't reflect a permanent, work-related shift. The case gets re-evaluated and comes off the log. This is exactly the kind of judgement call the rule expects a qualified professional to make — not a spreadsheet.

One EHS manager we spoke with put it this way: "The number on the printout is never the whole story. I've learned to wait for the audiologist's read before I even open the log, because half the borderline ones flip once someone who actually knows ears looks at them."

Why does age correction exist, and how does it change a determination?

Hearing naturally declines with age, independent of noise exposure. OSHA allows — but does not require — employers to adjust the current audiogram using published age-correction tables before comparing it to baseline. The tables subtract out the amount of hearing loss statistically expected from ageing alone, based on the employee's age and sex at the time of each test.

This is optional, not mandatory. An employer can choose to apply age correction or not, and that choice is made consistently, not selectively per employee to engineer an outcome. Because age correction changes the adjusted numbers, it can move a case in either direction. Applying it can turn a borderline recordable case into a non-recordable one, if a meaningful chunk of the shift is attributable to age rather than noise. It can also, less intuitively, occasionally tip a case toward recordable if the correction affects the baseline and current test unevenly. The direction depends on the individual numbers, not on a general rule of thumb.

Age correction only ever applies to comparisons against the STS threshold and the underlying hearing-level figures — it doesn't change the definition of the test itself, and it doesn't substitute for a physician's judgement about work-relatedness. For a fully worked scenario that walks through a hearing-loss determination end to end, see our hearing loss found on an annual audiogram scenario.

Where teams get this wrong

Hearing loss recording stays one of the more technically disputed categories in day-to-day EHS practice, and most of the disputes trace back to a handful of repeated mistakes.

• Recording every STS regardless of the 25 dB threshold. This is the single most common error. A shift alone triggers a closer look, not an automatic entry on the 300 Log. Skipping the absolute threshold check inflates the log with cases the rule doesn't require.
• Treating age correction as mandatory. It's optional. Some employers apply it inconsistently — using it only when it happens to remove a case from the log — which isn't how the standard is meant to work. The choice should be a documented, consistent policy, not a case-by-case decision aimed at a preferred outcome.
• Comparing against the wrong baseline. A baseline audiogram can be revised — for example, after a documented STS, OSHA permits the employer to substitute the STS-triggering audiogram as the new baseline going forward. Teams that keep comparing every future test against the original hire-date baseline, after a revision should have occurred, will keep flagging shifts that a correctly updated baseline wouldn't show.
• Skipping the physician or audiologist review. The rule builds in professional judgement on work-relatedness and permanence. A determination made purely from the raw dB numbers, without that review, skips a step the standard actually requires.
• Not re-evaluating after a retest. If a retest shows a shift has reversed — often because the original test caught a temporary condition like an ear infection or wax build-up — the case can and should be updated. Some logs never get revisited once the first entry is made.

For the broader mechanics of how any of these binary tests interact with the general recording rule, 1904.7 is worth reading alongside 1904.10 — it's the general criteria section that 1904.10 sits underneath. And if you're unsure whether a specific determination is right, OSHA's own Letters of Interpretation cover several real hearing-loss edge cases employers have asked about directly.

Where this fits in your recordkeeping

Once a hearing loss case is confirmed recordable, it goes on the 300 Log like any other case, and it's automatically classified as a case involving days away, restricted work, or none of those, depending on the treatment path — hearing loss cases are also, by definition, always at least "other recordable," per 1904.10(a). Getting the initial determination right matters, because it flows straight through to your 300A summary totals and, if your establishment is covered, your ITA electronic submission.

Job13 runs this exact two-part test — the STS threshold and the 25 dB absolute floor — against the numbers you enter, and returns a clear recordable or not-recordable answer. Where the professional judgement calls in 1904.10 genuinely can't be reduced to a formula — work-relatedness of an STS, permanence after a retest — it returns "needs review" instead of guessing. Run a hearing loss case through the free check at /recordable and see the reasoning behind the answer, not just the answer.

This isn't legal advice. For a borderline or disputed determination, talk to an audiologist, occupational physician, or OSHA compliance specialist.

Frequently asked questions

Does a standard threshold shift always mean the case is OSHA recordable?

No. An STS is only half the test. The case is recordable only if the employee's total hearing level, averaged at 2000, 3000, and 4000 Hz, is also 25 dB or more above audiometric zero in the same ear. A confirmed 10 dB or greater shift with a current average under 25 dB is not recordable under 1904.10.

Is age correction required when evaluating a hearing loss case?

No, age correction is optional under 1904.10. Employers may adjust the most recent audiogram using age-correction tables before comparing it to baseline, but they aren't required to. The choice should be applied consistently rather than selectively.

What happens if a retest shows the hearing loss has reversed?

The case can be re-evaluated. If a physician or audiologist determines the original shift wasn't permanent — for example, it was caused by a temporary condition like fluid in the ear — the case record can be updated to reflect that finding, including removing it from the log if it was already recorded.

Which audiogram counts as the baseline for comparison?

Normally the baseline is the audiogram taken at or near hire, before significant occupational noise exposure. Under the hearing conservation standard, the baseline can be revised — for instance, an audiogram that triggered an STS can become the new baseline for future comparisons, which is why using an outdated original baseline is a common source of errors.

Does 1904.10 apply the same way across every industry?

The recording criteria in 1904.10 apply the same way to any covered employer under 29 CFR Part 1904, regardless of industry, once a work-related hearing loss case meets both parts of the test. Job13 covers federal Part 1904 recording determinations; it doesn't currently address state-plan variations.