Hearing Protector Classes in Canada: What CSA Z94.2 Requires Beyond the NRR (2026)

Canada does not select hearing protection by the NRR on the box. CSA Z94.2 assigns every hearing protector a class, A, B or C, and the class tells you the range of workplace noise levels the device is suitable for. NRR is a supporting number that has to be derated before it means anything about your actual exposure.

This matters because almost every hearing protector sold in Canada is marketed on its NRR, a figure derived from an American EPA scheme built on a test method withdrawn decades ago. The number is not wrong. It is just not the thing your provincial regulator is asking about.

The A, B, C system is the Canadian one

Under the class system, a protector is assigned a single letter using laboratory attenuation data. The letter does not describe the sound level reaching your ear. It indicates the range of noise levels the protector should be used within to avoid the risk of hearing loss, which is a different and more directly useful thing.

Classes are pre-assigned according to defined attenuation ranges, measured in octave bands. In practice you will also see the variants A/AL, B/BL and C, where the L designations relate to low-frequency performance, which matters where the noise source is dominated by low frequencies such as large diesel engines and compressors.

Using it is straightforward in principle: measure the worker's noise exposure in dBA, then read across the standard's selection table to the class required at that exposure. Ontario's guide to its noise regulation describes the same approach, noting that CSA Z94.2-14 pre-assigns devices into classes according to defined attenuation ranges.

The current edition is CSA Z94.2-14, reaffirmed in 2024. It is the seventh edition, superseding versions published in 2002, 1994, 1984, 1979, 1974 and 1965.

What NRR actually is, and why it needs derating

The seventh edition of Z94.2 did add NRR, alongside an applicable derating scheme. That pairing is the important part, and it is the part that gets dropped.

CCOHS guidance on hearing protectors quotes the standard's definition in full, and it is worth noticing what sits inside it: the NRR is computed using mean attenuation data less two standard deviations, derived from the experimenter-fit procedure of the 1974 edition of ANSI Z3.19, a standard that has since been withdrawn, as defined in a 1979 EPA scheme.

Two things follow. First, the figure comes from a laboratory procedure where trained experimenters fitted the protectors, which is not how a worker inserts a foam plug at 6am. Second, much of the attenuation data still in use today, including the NRRs printed on packaging, comes from that 1974 method.

That gap between laboratory and job site is why derating exists. Taking the printed NRR at face value overestimates the protection a worker is actually getting, sometimes substantially. Your provincial regulator will specify which derating approach applies; the number on the box is the starting point of the calculation, not the end of it.

A rating is a group average, not a promise to you

This is the single most useful thing to understand about every number in this category.

Class, NRR and SNR are all obtained by calculation from tests on many subjects, and Canadian acoustics commentary is blunt that they should not be applied to individuals. An NRR of 20 does not mean every user gets 20 dB of attenuation from that protector. It means the population tested averaged a certain performance under laboratory fitting conditions.

Head shape, ear canal geometry, insertion technique, glasses, hair and hard hat interaction all move the real number, and they move it for each person differently. Two workers on the same crew wearing identical plugs can be receiving very different protection.

The seventh edition addresses this with Field Attenuation Estimation Systems, covered in its own section. FAES devices measure the attenuation actually delivered to an actual user, producing a personal attenuation rating rather than a group figure. Ontario's guidance notes the trend toward using both individual fit testing and group attenuation together for a more conservative estimate.

If that sounds familiar, it should. It is the same logic as respirator fit testing, where a correctly rated filter behind a poor seal protects nobody, covered in our guide to respirator fit testing under CSA Z94.4.

Above 105 dBA, one device is not enough

There is a hard threshold worth knowing. Where noise exposure exceeds an eight-hour equivalent level of 105 dBA, dual protection is called for: earplugs and earmuffs worn together, with a minimum of a Class B earmuff combined with a Class A earplug.

The standard also requires octave-band noise measurements at exposures above 105 dBA rather than a simple dBA reading, because at that level the frequency content of the noise changes which protector actually works.

Dual protection does not simply add the two ratings together, which is the intuitive assumption and the wrong one. The combination provides less than the sum, and the standard sets out how to account for it. If your workplace is running at these levels, this is a measurement and engineering problem before it is a purchasing decision.

Selecting for the job, not for the biggest number

Overprotection is a genuine hazard rather than a cautious choice. A worker who cannot hear a reversing alarm, a shouted warning or an equipment fault has traded one risk for another, and workers who feel cut off tend to break the seal or remove the device entirely.

Match the class to the measured exposure. Then choose the form factor for the work.

Earmuffs give consistent protection and suit intermittent removal, since they go on and off in a second. The Sellstrom over-the-head earmuffs at NRR 31 dB are the higher-attenuation option in that style.

Where a hard hat is mandatory, cap-mounted muffs avoid the conflict between a headband and a suspension, which is one of the most common causes of a broken seal. The Sellstrom cap-mounted earmuffs at NRR 28 dB attach directly to a hard hat, and pair with options such as the Pyramex Ridgeline cap style hard hat. Our hard hat safety guide covers head protection selection.

Banded plugs suit work where protection comes on and off repeatedly through a shift and muffs would be cumbersome. The Sellstrom premium banded ear plugs hang around the neck between uses, which is what stops them being left in a toolbox.

Confirm the class of any specific device against your measured exposure and your provincial requirement before purchase. Attenuation figures are published per model, and a form factor preference should never override the class the exposure calls for.

Compatibility is where protection quietly fails

Hearing protection is worn alongside everything else on the head, and each item is a candidate for breaking the seal.

Safety glasses are the most common culprit. A temple arm running under an earmuff cushion lifts the seal and opens a leak path that no attenuation rating accounts for. Thin, flat temple arms exist specifically for this, and our safety glasses guide covers eye protection selection.

Winter headwear does the same thing. A toque under muff cushions puts fabric between the cushion and the skin, and the thicker the knit the worse the leak. Our guide to cold stress at work in Canada covers the layering side of that trade-off.

Long hair, jewellery and respirator straps all belong in the same category. Where a tight-fitting respirator and hearing protection are both required, check the combination in practice rather than assuming both perform as rated.

Certified, or claimed?

One detail deserves attention because it is unusual among the CSA standards covering PPE.

When the seventh edition was published, Canadian acoustics commentary noted that no process for certification was then in place at CSA for confirming that a hearing protector actually meets the Z94.2 requirements. Verify the current position with your supplier rather than assuming a certification mark exists in the way it does for footwear.

The practical implication is that a claim of conformance to Z94.2 is a manufacturer's statement supported by test data, so ask for the attenuation figures and the class rather than accepting the standard number as a badge. That distinction between certified and compliant runs through the whole PPE category, covered in our guide to CSA certified versus CSA compliant.

Note also that Z94.2 tells users to refer to applicable local occupational health and safety regulations, which can require additional or superior performance. The standard is a floor, and your province may sit above it. For the wider picture see our hearing protection guide for industrial workers, the complete PPE checklist for Canadian workers, and our guide to PPE requirements for Canadian construction sites.

Frequently Asked Questions

What do hearing protector Classes A, B and C mean in Canada?

Under CSA Z94.2, each hearing protector is assigned a class based on its measured octave-band attenuation. The letter indicates the range of workplace noise levels the device should be used within to avoid the risk of hearing loss, rather than the sound level reaching the ear. Variants such as AL and BL relate to low-frequency performance.

Is NRR the right measure to use in Canada?

Not on its own. The seventh edition of Z94.2 includes NRR together with an applicable derating scheme, and the Canadian selection system is based on classes. NRR is derived from a 1979 EPA scheme using a since-withdrawn 1974 ANSI test method, so it needs derating before it describes real-world protection.

Does an NRR of 25 mean I get 25 dB of protection?

No. Class, NRR and SNR figures are calculated from tests across many subjects and should not be applied to individuals. Fit, head shape, insertion technique and other worn equipment all change the actual attenuation. Field attenuation estimation systems exist to measure the protection a specific worker actually receives.

When is double hearing protection required?

Where noise exposure exceeds an eight-hour equivalent level of 105 dBA, dual protection is called for, with a minimum of a Class B earmuff worn with a Class A earplug. Octave-band noise measurements are also required above that level rather than a simple dBA reading.

Can I just buy the highest NRR available?

That is usually a mistake. Overprotection can prevent a worker hearing reversing alarms, verbal warnings and equipment faults, and workers who feel cut off tend to break the seal or remove the device. Match the class to the measured exposure rather than maximizing the number.

Which edition of CSA Z94.2 is current?

Z94.2-14, reaffirmed in 2024. It is the seventh edition, superseding editions published in 2002, 1994, 1984, 1979, 1974 and 1965. It expanded the rating schemes, added NRR with a derating scheme, and introduced coverage of field attenuation estimation systems.

This guide is general workplace safety information, not legal or audiological advice. Hearing protection selection depends on measured noise exposure, and requirements are set by provincial and territorial regulation. Confirm yours with the occupational health and safety authority in your jurisdiction.