Putting earmuffs over earplugs earns just 5 dB of extra credit under the US estimating rule, no matter how high the two labels add up (OSHA Technical Manual, Section III Chapter 5, Appendix F). NIOSH’s own criteria document says double protection adds only 5 to 10 dB, and warns that even two protectors are inadequate once 8-hour exposure passes 105 dBA (NIOSH Criteria Document 98-126, 1998). The reason is anatomical: sound keeps reaching the inner ear through the skull, and mean bone-conduction limits sit between 40 and 60 dB across 125 Hz to 8 kHz (Berger, Kieper and Gauger, JASA 2003). Impulse noise is the exception, with double protection adding 15 to 20 dB against 159 to 169 dB gunfire at a NIOSH-studied firing range (NIOSH Publication 2013-104). We aggregated data from OSHA, NIOSH, MSHA, the US EPA, the UK HSE, peer-reviewed acoustics journals and military medicine research; for the broader picture see our hearing protection statistics, and for single-device fit results our hearing protection fit-testing and PAR statistics.
TL;DR
- OSHA adds 5 dB to the higher NRR after a 7 dB dBA correction, so 110 dBA with an NRR 29 plug and NRR 25 muff becomes 83 dBA (OSHA Technical Manual, Appendix F).
- NIOSH recommends double protection above a 100 dBA 8-hour TWA, says it adds only 5 to 10 dB, and calls it inadequate above 105 dBA (NIOSH Criteria Document 98-126).
- MSHA’s Dual Hearing Protection Level is a TWA8 of 105 dBA, a dose of 800% (30 CFR 62.101).
- Mean bone-conduction limits to attenuation run from 40 to 60 dB between 125 Hz and 8 kHz (Berger et al., JASA 2003).
- The 1.5 to 2 kHz bone-conduction peak caps mean mid-frequency attenuation at about 40 dB (Homma et al., JASA 2009).
- Combining plugs and muffs gained 3 to 10 dB of NRR over the better single device, and never more than 3 dB at 2 kHz (3M EARLog 13).
- Against a 168 dB impulse, a plug plus muff cut the peak by 46.3 dB versus 37.8 dB for the muff alone (Murphy et al., Noise and Health 2015).
- Aircraft engine run-ups averaged 107.7 dBA Leq with peaks averaging 143.4 dBA (Aljaroudi, AHFE 2026).
- Hearing protectors roughly doubled front-back confusions and made localization about 50% worse (Esparza et al., Military Medicine 2026).
1. The 5 dB Rule: How OSHA and NIOSH Credit Dual Protection
The US system treats the second protector as a modest bonus, not a second wall. OSHA credits only 5 dB for adding earmuffs over earplugs, applied to whichever device has the higher Noise Reduction Rating (NRR, the laboratory attenuation number printed on the package). The arithmetic is deliberately conservative: a 29 dB plug and a 25 dB muff do not make 54 dB, they make 27 dB of usable credit once the 7 dB dBA correction and the 5 dB bonus are applied: (29 - 7) + 5 = 27.
The two big US agencies agree on direction but not on the exact figure. OSHA’s enforcement formula uses a flat 5 dB; NIOSH’s 1998 criteria document says double protection adds 5 to 10 dB. When OSHA’s 50% field safety factor is layered on top, the same worker who looks well protected on paper (83 dBA) lands at 94 dBA, still above OSHA’s 90 dBA permissible level, which is the signal OSHA uses to decide that engineering controls are needed.
| Metric | Value | Source |
|---|---|---|
| Credit added for wearing earmuffs over earplugs (dual protection) | 5 dB added to the higher of the two NRRs, after subtracting 7 dB when exposure is measured in dBA | OSHA Technical Manual, Section III Chapter 5, Appendix F |
| OSHA worked example: 110 dBA TWA, NRR 29 earplug plus NRR 25 earmuff | 83 dBA estimated exposure, from 110 - [(29 - 7) + 5] | OSHA Technical Manual, Appendix F |
| Same worked example with OSHA’s 50% workplace safety factor applied to the higher NRR | 94 dBA (our calculation: 110 - {[(29 - 7) x 50%] + 5}) | OSHA Technical Manual, Appendix F formula |
| OSHA single-protector example: 100 dBA TWA, NRR 19 earmuff only | 88 dBA estimated exposure | OSHA Technical Manual, Appendix F |
| Extra attenuation that double protection adds, per NIOSH | only 5 to 10 dB | NIOSH Criteria Document 98-126 (1998), Chapter 6 |
| Level hearing protectors must reduce exposure to: workers without a threshold shift vs workers with a standard threshold shift | 8-hour TWA of 90 dBA vs 85 dBA or below | OSHA 29 CFR 1910.95(j) |
Context note: TWA means time-weighted average, the noise level averaged over a work shift. OSHA states that the 50% safety factor is used when judging whether engineering controls are needed, while hearing conservation enforcement uses only the 7 dB subtraction.
2. When Dual Protection Kicks In: Exposure Thresholds
The trigger levels show how narrow the window for dual protection really is. NIOSH recommends double protection above a 100 dBA 8-hour TWA, yet calls even double protection inadequate above 105 dBA, a band only 5 dB wide. That is the same 5 dB the OSHA credit in Section 1 grants, so on paper the second protector buys about 5 dB of headroom and no more.
Mining regulation draws the line in the same place. MSHA’s Dual Hearing Protection Level sits at a TWA8 of 105 dBA, which it expresses as a dose of 800%, eight times the daily noise allowance, and it bans any sound level above 115 dBA measured without credit for protectors (see our underground mining noise exposure statistics for where those levels occur). In the UK, the HSE guide stops at the 100 to 105 dB band, where it suggests protectors rated at SNR 30 or more.
| Metric | Value | Source |
|---|---|---|
| 8-hour TWA above which NIOSH says workers should wear double hearing protection | 100 dBA | NIOSH Criteria Document 98-126, Section 1.5 |
| TWA above which NIOSH says even double protection is inadequate | 105 dBA | NIOSH 98-126, Chapter 6 |
| MSHA Dual Hearing Protection Level (US mining) | TWA8 of 105 dBA, a dose of 800% | 30 CFR 62.101 |
| MSHA requirement above the dual level | concurrent use of an earplug and an earmuff | 30 CFR 62.140 |
| MSHA ceiling on any sound level, without adjustment for hearing protectors | 115 dBA | 30 CFR 62.130(c) |
| MSHA permissible exposure level vs action level | 90 dBA (100% dose) vs 85 dBA (50% dose) | 30 CFR 62.101 |
| OSHA Table G-16 permissible daily durations at 105, 110 and 115 dBA | 1 hour, half an hour, a quarter hour or less | OSHA 29 CFR 1910.95(b) |
| HSE suggested protector rating for 100 to 105 dB task noise | SNR of 30 or more | UK HSE, Hearing protection |
Context note: the NIOSH recommended exposure limit (REL) is 85 dBA as an 8-hour TWA with a 3 dB exchange rate (every 3 dB increase halves the allowed exposure time), and NIOSH 98-126 estimates an 8% excess risk of hearing loss for workers exposed to an average daily level of 85 dBA.
3. What the Lab Measures When Plugs and Muffs Are Combined
The most detailed public dataset on combinations is still Elliott Berger’s EARLog 13, and its central finding is that the bonus depends on the earplug far more than on the muff. Combining a plug and a muff gained only 3 to 10 dB of NRR over the better single device, while the gain over plugs alone ranged from 7 to 17 dB. Weak plugs gain the most from a muff; a deeply inserted foam plug gains the least, because it is already close to the ceiling described in Section 4.
The frequency picture is starker. At individual frequencies the gain over the better device varied from approximately 0 to 15 dB, but at 2 kHz no combination gained more than 3 dB. Berger concluded that no easy rule of thumb could predict combined attenuation from single-device results, which is why the OSHA 5 dB figure is a policy convention rather than a measurement.
| Metric | Value | Source |
|---|---|---|
| Combined NRR: deep-insertion foam earplug (NRR 31 alone) plus earmuff #2 (NRR 21 alone) | 35 | 3M EARLog 13, Berger |
| Combined NRR: premolded earplug (NRR 14 alone) plus earmuff #2 | 29 | 3M EARLog 13 |
| Combined NRR: fiber earplug (NRR 8 alone) plus earmuff #2 | 25 | 3M EARLog 13 |
| NRR gain of the combination over the better single device | 3 to 10 dB | 3M EARLog 13 |
| NRR gain of the combination over earplugs alone vs over earmuffs alone | 7 to 17 dB vs 3 to 14 dB | 3M EARLog 13 |
| Gain over the better single device at individual test frequencies | approximately 0 to 15 dB, never more than 3 dB at 2 kHz | 3M EARLog 13 |
Most recent available data: 3M EARLog 13 (first printed 1984, reprinted 1999). No newer open dataset of combined NRRs was reachable for this roundup; a 2022 Applied Acoustics paper by Trompette and colleagues revisits the question but its figures were not publicly accessible.
4. The Bone-Conduction Ceiling
Bone conduction means sound that travels through the bones of the skull and the tissues of the head straight to the inner ear, bypassing the ear canal that plugs and muffs block. Mean bone-conduction limits to protector attenuation run from 40 to 60 dB between 125 Hz and 8 kHz, which is why stacking a 33 dB plug with a 28 dB muff cannot yield anything like 61 dB. The weakest point is in the speech range: a bone-conduction peak near 1.5 to 2 kHz, which researchers link to a resonance of the middle ear, holds mean attenuation to about 40 dB there.
The ceiling is not absolute. Berger’s 2003 study showed gains exceeding 10 dB above the head-not-covered limits when the head itself is shielded, the logic behind flight helmets. Laboratory work also traced part of the double-protector shortfall to sound entering through the earmuff cushion and travelling through the earplug and ear canal walls, a structure-borne path that dominates above 300 Hz (Luan et al., Applied Acoustics 2021).
| Metric | Value | Source |
|---|---|---|
| Mean bone-conduction limits to hearing protector attenuation, 125 Hz to 8 kHz | 40 to 60 dB | Berger, Kieper and Gauger, JASA 2003 |
| Extra attenuation above the head-not-covered limit when the head is shielded from sound | more than 10 dB | Berger, Kieper and Gauger, JASA 2003 |
| Mean attenuation limit in the critical mid-frequency range, set by the 1.5 to 2 kHz bone-conduction peak | about 40 dB | Homma et al., JASA 2009 |
| Location of the bone-conduction limit peak at higher frequency resolution | around 1.6 to 1.7 kHz | Homma et al., JASA 2009 |
| Lowest bone-conduction limit in the impulse-testing standard | 41 dB at 2000 Hz | Murphy et al., Noise and Health 2015 |
| Frequency above which the structure-borne path through the earmuff cushion dominates the double-protector shortfall | 300 Hz | Luan et al., Applied Acoustics 2021 |
| Cut in bone-conducted stapes motion from negative ear-canal pressure of -400 mm H2O (cadaver bones) | about 5 to 8 dB across 0.8 to 2.5 kHz; peak 8.6 (plus or minus 4.7) dB at 1.6 kHz | Homma et al., Hearing Research 2010 |
Outlier note: the 2010 pressurization result comes from human cadaver temporal bones, not living wearers, so it describes a mechanism, not a product. Most recent available data on the core limits: Berger et al., 2003, and Homma et al., 2009.
5. Dual Protection on Real Ears: The 2025 and 2026 Data
Newer studies confirm the ceiling on living listeners. A 2025 real-ear study of an Otoblock-style earplug under an earmuff measured peak attenuation of about 36 dB near 3000 Hz with a probe microphone, and about 38 dB near 4000 Hz by behavioral hearing thresholds (Hauptman et al., Forum Acusticum 2025). Both peaks sit just below the bottom of the 40 to 60 dB bone-conduction band from Section 4, so two sealed layers end up close to the ceiling, not past it.
The two methods also disagree in a revealing way. Probe-microphone measurements tend to overestimate low-frequency attenuation under double protection, because the microphone sits in the sealed canal and cannot hear what reaches the cochlea through bone. Researchers who deliberately simulate one-sided hearing loss with a 33 dB plug under a 28 dB muff note the same leak: at a 70 dB SPL presentation level, some sound may still reach the plugged ear by bone conduction.
| Metric | Value | Source |
|---|---|---|
| Participants in the 2025 double-protection real-ear study (Phonak EasyView Otoblock earplug plus 3M Optime 105 earmuff) | 10, aged 18 to 28 | Hauptman et al., Forum Acusticum 2025 |
| Peak attenuation measured with a probe microphone in the ear canal | about 36 dB near 3000 Hz | Hauptman et al., Forum Acusticum 2025 |
| Lowest probe-microphone attenuation in the tested range | 18 dB at 12,000 Hz | Hauptman et al., Forum Acusticum 2025 |
| Peak attenuation measured by behavioral hearing thresholds | about 38 dB near 4000 Hz | Hauptman et al., Forum Acusticum 2025 |
| Behavioral attenuation at the upper end of the tested range | 33 to 40 dB | Hauptman et al., Forum Acusticum 2025 |
| Labeled NRRs of the foam earplug and earmuff used to simulate one-sided hearing loss | 33 dB and 28 dB, in 16 normal-hearing participants | Epperson et al., The Laryngoscope 2025 |
Context note: few fit-test systems can measure earmuffs, and fewer still can estimate double protection, according to a 2026 Walter Reed validation study of a neck-based field attenuation system (Voelker et al., JASA 2026). That gap is why combined protection still leans on the OSHA formula rather than on individual measurement.
6. Impulse Noise: Firearms and Engine Run-Ups
Impulse noise, short bursts such as gunshots or blasts, is where a second protector earns its keep. On an acoustic test fixture, a plug and muff combination cut a 168 dB impulse by 46.3 dB, versus 37.8 dB for the earmuff alone and 29.0 dB for the earplug alone (Murphy et al., Noise and Health 2015). The edge of the combination over the muff alone was 8.2, 9.0 and 8.5 dB at the 134, 150 and 168 dB levels (our subtraction from the table below), and NIOSH field work at a firing range found double protection added 15 to 20 dB.
These are the settings where single protection runs out. A 2026 field measurement of aircraft engine run-ups (ground tests of engines at power) found every sample above 126 dBA maximum and peaks averaging 143.4 dBA, above the 140 dBA ACGIH ceiling for impulse noise; the author concluded single protection was insufficient (see our aviation ramp and ground crew hearing protection statistics). The human cost shows up in a French military review of 419 acute acoustic trauma cases, where the 5.56 mm assault rifle caused 78.0% of injuries and tinnitus affected 87.4% (Lachaux et al., Military Medicine 2024).
| Metric | Value | Source |
|---|---|---|
| Impulse peak insertion loss (peak reduction at the fixture’s ear) at 134 dB: earplug plus earmuff vs earmuff alone | 30.9 dB vs 22.7 dB | Murphy et al., Noise and Health 2015 (E-A-RCAL fixture) |
| Impulse peak insertion loss at 150 dB: earplug plus earmuff vs earmuff alone vs earplug alone | 38.0 dB vs 29.0 dB vs 19.5 dB | Murphy et al., Noise and Health 2015 |
| Impulse peak insertion loss at 168 dB: earplug plus earmuff vs earmuff alone | 46.3 dB vs 37.8 dB | Murphy et al., Noise and Health 2015 |
| Firearm noise measured at a SWAT training site | 159 to 169 dB peak | NIOSH Publication 2013-104 |
| Peak noise reduction from single earplugs, earmuffs or custom protectors at the range | in the 30 dB range | NIOSH Publication 2013-104 |
| Additional protection from double protection (plugs plus muffs) at the range | 15 to 20 dB | NIOSH Publication 2013-104 |
| Aircraft engine run-up equivalent continuous level (Leq) | 105.8 to 109.6 dBA, mean 107.7 dBA | Aljaroudi, AHFE 2026 |
| Aircraft engine run-up peak level | average 143.4 dBA, maximum above 126 dBA in all samples | Aljaroudi, AHFE 2026 |
Context note: in the Lachaux review, more than 20% of patients still had long-term hearing loss on late audiograms taken a mean 448.0 days after the injury.
7. Labels, Ratings and the Cost of Overprotection
The NRR on the box was never designed to describe a combination. The current US method, unchanged since 1979, assumes 100 dB pink noise in seven octave bands and subtracts two standard deviations of the test panel’s attenuation plus a 3 dB spectral uncertainty term, giving a single number for a single device (40 CFR 211.207). EPA’s 2009 proposal, written when 1,029 products were on sale in the US, would have replaced it with a range: a lower rating met by at least 80% of test subjects and a higher one met by at least 20%. But no rating in either system tells a buyer what a plug and muff deliver together.
Stacking protectors also costs information. The UK HSE advises avoiding protectors that bring the level at the ear below 70 dB, because overprotection hides speech and warning signals. A 2026 randomized trial found hearing protectors roughly doubled front-back confusions and made sound localization about 50% worse, though brief training recovered much of the loss (Esparza et al., Military Medicine 2026).
| Metric | Value | Source |
|---|---|---|
| Noise assumed in each octave band when computing the US NRR | 100 dB pink noise across 7 bands, 125 to 4000 Hz | 40 CFR 211.207 |
| Margins subtracted in the US NRR computation | 2 standard deviations of attenuation plus a 3 dB spectral uncertainty | 40 CFR 211.207 |
| EPA’s own worked NRR example | 19.8 dB, labeled as NRR 20 | 40 CFR 211.207 |
| 2009 EPA proposed range: lower rating vs higher rating | protection reached by at least 80% vs at least 20% of test subjects | EPA proposed rule, Federal Register 2009 |
| Hearing protection products on sale in the US in the 2009 EPA count | 1,029 (403 earplug or semi-aural, 572 earmuff, 54 active noise reduction) | EPA proposed rule, Federal Register 2009 |
| Level at the ear below which the HSE says protectors should be avoided | 70 dB | UK HSE, Hearing protection |
| Effect of hearing protectors on front-back confusions vs open ears, Day 1 | roughly doubled, with localization about 50% worse | Esparza et al., Military Medicine 2026 |
Context note: the Esparza trial randomized 24 participants to training and 23 to a control group; the abstract does not say whether the protectors were earplugs, earmuffs or both.
Summary: Dual Hearing Protection by the Numbers
| Metric | Value | Source |
|---|---|---|
| OSHA credit for adding earmuffs over earplugs | 5 dB on the higher NRR, after a 7 dB dBA correction | OSHA Technical Manual, Appendix F |
| OSHA example: 110 dBA, NRR 29 plug plus NRR 25 muff, without vs with the 50% safety factor | 83 dBA vs 94 dBA (second value our calculation) | OSHA Technical Manual, Appendix F |
| Attenuation added by double protection, per NIOSH | 5 to 10 dB | NIOSH 98-126 |
| NIOSH TWA for recommending double protection vs TWA where it becomes inadequate | above 100 dBA vs above 105 dBA | NIOSH 98-126 |
| MSHA Dual Hearing Protection Level | 105 dBA TWA8, 800% dose | 30 CFR 62.101 |
| MSHA ceiling without protector credit | 115 dBA | 30 CFR 62.130 |
| NRR gain of plug plus muff over the better single device, overall vs at 2 kHz | 3 to 10 dB vs no more than 3 dB | 3M EARLog 13 |
| Mean bone-conduction limits, 125 Hz to 8 kHz | 40 to 60 dB | Berger et al., JASA 2003 |
| Mid-frequency attenuation limit from the 1.5 to 2 kHz peak | about 40 dB | Homma et al., JASA 2009 |
| Real-ear double protection peak, probe microphone vs behavioral | about 36 dB vs about 38 dB | Hauptman et al., Forum Acusticum 2025 |
| Impulse reduction at 168 dB, plug plus muff vs muff alone | 46.3 dB vs 37.8 dB | Murphy et al., Noise and Health 2015 |
| Firearm noise at a SWAT training site | 159 to 169 dB peak | NIOSH Publication 2013-104 |
| Added protection from double protection at a firing range | 15 to 20 dB | NIOSH Publication 2013-104 |
| Aircraft engine run-up mean Leq vs average peak | 107.7 dBA vs 143.4 dBA | Aljaroudi, AHFE 2026 |
| Acute acoustic trauma cases caused by the 5.56 mm assault rifle (French military, 419 patients) | 78.0% | Lachaux et al., Military Medicine 2024 |
| Level at the ear below which the HSE advises against protectors | 70 dB | UK HSE, Hearing protection |
| Effect of hearing protectors on front-back confusions | roughly doubled | Esparza et al., Military Medicine 2026 |
Methodology and Sources
- OSHA Technical Manual, Section III Chapter 5 (Noise), including Appendix F on NRR methods
- OSHA 29 CFR 1910.95, Occupational Noise Exposure
- NIOSH Criteria for a Recommended Standard: Occupational Noise Exposure, Revised Criteria 1998 (Publication 98-126)
- NIOSH Reducing Exposure to Lead and Noise at Outdoor Firing Ranges (Publication 2013-104)
- MSHA occupational noise rule: 30 CFR 62.101, 30 CFR 62.130, 30 CFR 62.140
- US EPA NRR computation, 40 CFR 211.207
- US EPA proposed rule on hearing protection device labeling, Federal Register, August 5, 2009
- UK Health and Safety Executive, Hearing protection
- Berger EH, 3M EARLog 13: Attenuation of Earplugs Worn in Combination with Earmuffs
- Berger EH, Kieper RW, Gauger D. Hearing protection: surpassing the limits to attenuation imposed by the bone-conduction pathways. JASA 2003
- Homma K et al. Ossicular resonance modes of the human middle ear for bone and air conduction. JASA 2009
- Homma K et al. Effects of ear-canal pressurization on middle-ear bone- and air-conduction responses. Hearing Research 2010
- Luan Y et al. Experimental study of earplug noise reduction of a double hearing protector on an acoustic test fixture. Applied Acoustics 2021
- Murphy WJ et al. Measurement of impulse peak insertion loss from two acoustic test fixtures and four hearing protector conditions. Noise and Health 2015
- Hauptman A et al. Sound attenuation with double hearing protection. Forum Acusticum 2025
- Epperson MV et al. The Laryngoscope 2025 (unilateral hearing loss simulated with earplug plus earmuff)
- Aljaroudi A. Occupational noise exposure during aircraft engine run-up task in aviation maintenance. AHFE 2026
- Lachaux J et al. Long-term hearing loss after acute acoustic trauma in the French military. Military Medicine 2024
- Esparza LV et al. Impact of brief training on spatial hearing and motion perception while wearing hearing protection. Military Medicine 2026
- Voelker J et al. Electroacoustic validation of a neck-based field attenuation estimation system. JASA 2026 (abstract only).
- Data watch: sources disagree on how much a second protector adds. OSHA uses a flat 5 dB, NIOSH 98-126 says 5 to 10 dB, EARLog 13 measured 3 to 10 dB of NRR over the better device, and NIOSH’s firing-range report found 15 to 20 dB against impulse peaks; the gap reflects continuous vs impulse noise and lab vs field methods, not an error. The Aljaroudi 2026 paper reports a Leq range of 64 to 127 dBA and a 144 dBA average peak in its abstract but 105.8 to 109.6 dBA and 143.4 dBA in its results; we cite the results section. Hauptman et al. place the crossover between methods near 3000 Hz in the results and near 4 kHz in the abstract. EARLog 13 (1984), Berger et al. (2003) and Homma et al. (2009) are the most recent available data for their measurements. The OSHA 50% safety-factor result of 94 dBA is our arithmetic using OSHA’s published formula. EPA’s 2009 proposal has not replaced the 1979 NRR method still codified at 40 CFR 211.207, and NIOSH has since replaced its 1998 derating guidance with a recommendation for individual fit testing, which does not yet cover dual protection in most systems.
Last updated: October 18, 2026. We update this roundup quarterly, and the next refresh is expected when EPA acts on its hearing protector labeling rule (40 CFR Part 211) or NIOSH publishes new guidance on dual protection.