Why we derate the NRR
The Noise Reduction Rating (NRR) printed on every hearing protector package is measured under ANSI S3.19 laboratory conditions: trained subjects, optimal insertion technique, controlled acoustic environment. In a real workplace, workers insert earplugs hastily, earmuffs are worn over glasses or hard-hat straps, and fit quality varies dramatically between individuals. The gap between laboratory and real-world attenuation is well documented in the scientific literature — and it is large.
Derating factors were developed to close this gap. The CCOHS method (CSA Z94.2-14) applies a 50% derating for earplugs (typically harder to fit correctly) and a 70% derating for earmuffs (easier to fit but still far below lab performance). Dual protection — wearing both earplugs and earmuffs simultaneously — uses a combined NRR formula with a 65% derating applied to the combined value, because the marginal gain of the second device diminishes rapidly beyond a certain attenuation level (the body transmits sound through bone and tissue, not just the ear canal).
The "−3 dB" correction applied when noise is measured in dBA (rather than dBC) accounts for the frequency-weighting offset between the A and C scales. The NRR is derived from octave-band measurements that better approximate C-weighting; a 3 dB correction is therefore applied when only the dBA value is available.
CCOHS vs OSHA vs NIOSH — three methods, three answers
Given the same NRR and noise level, these three methods produce different effective exposure estimates because they use different derating assumptions:
- CCOHS / CSA Z94.2-14 (this tool): 50% for earplugs, 70% for earmuffs, 65% for dual protection. Differentiates by device type, reflecting the observation that earmuffs are generally fit more consistently than foam earplugs. This is the standard referenced by the RSST (Règlement sur la santé et la sécurité du travail) in Quebec.
- OSHA (29 CFR 1910.95 Appendix B): 50% derating applied uniformly to all protector types. Simpler but less nuanced — it does not distinguish between earplugs and earmuffs, and it typically produces a more conservative (higher) effective level estimate than the CCOHS method for earmuffs.
- NIOSH 98-126 (historical, now superseded): applied a variable derating: 75% for slow-recovery foam earplugs, 50% for all other earplugs, 75% for earmuffs. This three-tier system was superseded in January 2025 by NIOSH 2025-104, which abandons the variable derating in favor of individual fit-testing (PAR).
What NIOSH 2025-104 changes
In January 2025, NIOSH published document 2025-104, which formally supersedes the 1998 document NIOSH 98-126. The key shift is conceptual: instead of applying a population-level derating factor to the manufacturer NRR, NIOSH now recommends that workplaces move toward individual fit-testing (PAR — Personal Attenuation Rating) for each worker. The rationale is that derating factors describe average populations and mask enormous individual variation — two workers wearing the same earplug model can experience 10–15 dB differences in actual attenuation.
A PAR test uses a portable fit-testing device (such as the 3M E-A-Rfit or Howard Leight VeriPRO) to measure the real attenuation achieved by a specific individual wearing a specific protector. This provides a Personal Attenuation Rating that reflects actual protection rather than a theoretical estimate. PAR testing is already recommended by ACGIH and is becoming the gold standard for mature hearing conservation programs.
Practical implication: for compliance and basic assessments, applying CCOHS derating factors (as this tool does) remains valid and is consistent with CSA Z94.2-14. For workers exposed near or above occupational exposure limits, or for high-attenuation requirements, individual fit-testing is strongly recommended.
Frequently Asked Questions
Should I use PAR or NRR derating?
What is the NRR (Noise Reduction Rating)?
Should I trust the manufacturer NRR label?
Does double protection always add up?
What is a fit-test (PAR — Personal Attenuation Rating)?
Which method should I use in Quebec?
You might also need
See all tools →Complementary tools based on what you're doing