How Loud Is a Fire Alarm? (Smoke Detector Decibel Levels)
A smoke alarm's rated loudness and the level that reaches a sleeping child are two different measurements, and the government's own field data shows how far apart they are.
Reviewed by our team against authoritative public-health sources. See our editorial standards.
Measurement context
The 85 dBA figure is a product requirement at the horn, not a level in any bedroom. Compliance is judged at the source; effectiveness happens after the sound has crossed a house.
Personal exposure
Not determined by these published figures. No safe listening time can be inferred.
Quick Answer
The requirement and the reality are different measurements. A listed smoke alarm must produce at least 85 dBA measured 10 feet from the horn under UL 217. But US Consumer Product Safety Commission field measurements in occupied test homes recorded 60–76 dBA in bedrooms with the door open, and 42–89 dBA with the door closed — closed doors cut 8 to 40 dBA. That matters because roughly 75 dBA is generally cited as needed to reliably wake adults, and one study found 85% of children did not reliably wake to a standard hallway alarm received at 60 dBA at the child's ear.
The 85 dBA figure is a product requirement at the horn, not a level in any bedroom. Compliance is judged at the source; effectiveness happens after the sound has crossed a house.
What 85 dBA actually refers to
A listed smoke alarm must produce not less than 85 dBA measured 10 feet from the alarm, under UL 217's test conditions. That is a manufacturing requirement, verified in a test room with nothing between the horn and the microphone.
It is a sensible way to write a product standard, because it is checkable and repeatable. It is a poor description of what happens in a house, and the agency responsible for consumer product safety has published the data proving it.
What the same alarms measure inside real homes
The CPSC measured sound levels in occupied test homes, directly under sounding alarms and in bedrooms with doors open and closed. Under the alarm, levels were 85–102 dBA, consistent with the requirement. In bedrooms with the door open, 60–76 dBA. With the door closed, 42–89 dBA depending on the house and the floor — closed doors reduced levels by 8 to 40 dBA.
That closed-door range is the crucial finding. It spans nearly 50 dB, which means the level reaching a sleeper is determined less by the alarm than by the building it is installed in.
| Position | Measured level |
|---|---|
| UL 217 requirement, at the horn | at least 85 dBA at 10 ft |
| Directly under a sounding alarm | 85–102 dBA |
| Bedroom, door open | 60–76 dBA |
| Bedroom, door closed | 42–89 dBA |
| Reduction attributable to a closed door | 8–40 dBA |
A small number of test houses across single and multi-storey layouts with standalone and interconnected alarms. Not a nationally representative sample.
Does it wake people?
CPSC's review of the arousal literature states that a minimum of about 75 dBA is required to wake most adults, although levels as low as 55–60 dBA can wake some. Compare that with the field data: a closed bedroom door can leave the level below 60 dBA, and in some measured houses far below.
For children the picture is worse. A Victoria University study found that 85% of children did not reliably wake to a standard hallway smoke alarm received at 60 dBA at the child's ear. Developmental arousal-threshold work cited in the CPSC review found that no participant under the age of 15 was woken by tones below an 80 dB average intensity, while the oldest subjects woke at an average of 50 dB.
This is the central, uncomfortable finding of the field: a fully compliant alarm can be inaudible enough, or simply ineffective enough, to fail to wake the person it is protecting.
| Population | Finding |
|---|---|
| Most adults | About 75 dBA required to wake reliably |
| Some adults | 55–60 dBA can be enough, depending on ambient noise |
| Children, standard tone alarm | 85% did not reliably wake at 60 dBA received at the ear |
| Participants under 15 (tone study) | None woken by tones below 80 dB average intensity |
| Oldest participants (same study) | Woken at 50 dB average intensity |
The child studies used standard high-pitch tone alarms and small experimental samples. The tone-threshold figures reach us through the CPSC review rather than the original papers.
This is not a contradiction in the sources
It would be easy to present the 85 dBA requirement and the 42 dBA bedroom measurement as a scandal or an error. Neither is true. They are the same alarms measured in different places, and the sources agree on the 85 dBA figure itself.
The mismatch is structural: product compliance is assessed at the source in an unobstructed test room, while wake-effectiveness research measures at the sleeper's ear after real attenuation. Both measurements are needed, and quoting only the first is what creates the false impression.
What we could not confirm
NFPA 72 contains a sleeping-area audibility requirement based on ambient noise plus a margin, measured at the pillow. We could only reach it through secondary references and could not read the operative text, so we are not quoting its numeric formula.
We could not confirm from primary text the sound output requirement in UL 2034 for carbon monoxide alarms, and we are explicitly not assuming it mirrors UL 217's 85 dBA at 10 feet. We also found no large modern nationally representative field study of at-pillow levels — the CPSC data comes from a small set of test houses from 2004 to 2007 — and no data on whether newer low-frequency 520 Hz alarms change the child wake-rate findings.
Checking your own alarm
This is one of the more useful things a browser meter can do, precisely because the question is comparative rather than absolute. Sound the alarm on test and compare the level under the alarm, in the hallway, in a bedroom with the door open, and with the door closed.
Uncalibrated, our meter reports A-weighted digital signal level in dBFS — relative, not sound pressure — which is entirely adequate for that comparison, since what you want is the difference between positions. After a one-time local calibration it reports an estimated dBA figure. What it cannot do is certify compliance with UL 217 or with any code requirement.
Hearing Safety
Evacuate promptly when an alarm sounds; brief exposure is not a guarantee of hearing safety. Never disable an alarm because of its noise. Have repeated false alarms investigated by a qualified person.
Frequently Asked Questions
How many decibels is a smoke alarm?
A listed alarm must produce at least 85 dBA measured 10 feet from the horn under UL 217. In practice, government field measurements found 85–102 dBA directly under a sounding alarm but only 60–76 dBA in bedrooms with the door open and 42–89 dBA with the door closed.
Will a smoke alarm wake me up?
It depends on the level reaching you. About 75 dBA is generally cited as needed to reliably wake most adults, and some wake at 55–60 dBA. Field measurements show a closed bedroom door can bring the level below that.
Do smoke alarms wake children?
Often not. One study found 85% of children did not reliably wake to a standard hallway alarm received at 60 dBA at the ear, and arousal research cited by the CPSC found no participant under 15 woken by tones below 80 dB average intensity.
How much difference does a closed bedroom door make?
In CPSC's test homes, 8 to 40 dBA — a very wide range that depends on the house, the floor and the layout rather than on the alarm.
Do carbon monoxide alarms have the same requirement?
We could not confirm the sound output clause in UL 2034 from primary text, and we are not assuming it matches UL 217's 85 dBA at 10 feet.
Where should I put alarms so they are actually heard?
The measurement point that matters is where the sleeper's head is, with the door in its normal night-time position. Interconnected alarms and an alarm inside the sleeping area address the attenuation the CPSC data documents; this page does not give installation code advice.
Explore More
Related Noise Guides
How Loud Is Snoring? Measured Levels and What They Indicate
Snoring loudness turns out to track a medical variable, which makes it one of the few sound levels with genuine diagnostic meaning.
7 min readMeasurement TipsHow Accurate Are Phone Decibel Meters?
Two research groups have tested phone sound meters properly — one at NIOSH, one across 100 handsets in a reverberation room. Their findings are more specific, and more useful, than "phones are inaccurate".
8 min readVehicle NoiseHow Loud Is a Police Siren? Decibel Levels Explained
Emergency sirens are built to a specification, not a habit. Three different standards define how loud one must be, at 3 metres, on axis — and they disagree by several decibels.
5 min readEvidence behind the levels on this page
Every sound level quoted above comes from one of the publications below. For each one we record what was actually measured, the microphone position or distance where that is stated, the metric used, and what the figure can and cannot tell you. Where a source did not state a condition, we say so rather than filling the gap.
The audibility of smoke alarms in residential homes, CPSC-ES-0503
US Consumer Product Safety Commission · 2005, revised 2007 · US federal agency field measurement report
- What was measured
- Sound levels measured in occupied test homes under sounding alarms and in bedrooms with doors open and closed
- Reported level
- Under the alarm 85–102 dBA; bedroom door open 60–76 dBA; bedroom door closed 42–89 dBA; closed doors reduced levels by 8–40 dBA. Quotes the UL 217 requirement of not less than 85 dBA at 10 ft
- Distance / position
- At the alarm and at bedroom positions including near the pillow, in multiple test houses
- Frequency weighting
- A
- Conditions
- Doors open and closed, single and multi-storey homes, standalone and interconnected alarms
- Supports on this page
- The field measurement table and the whole requirement-versus-reality argument
- Limitations
- A small number of test houses; not a nationally representative sample.
A review of the sound effectiveness of residential smoke alarms, CPSC-ES-0502
US Consumer Product Safety Commission · 2004, revised · US federal agency literature review
- What was measured
- Levels required to wake sleepers, and modelled attenuation from a sounding alarm through a house to a bedroom
- Reported level
- A minimum of about 75 dBA is required to wake most adults, though 55–60 dBA can wake some; modelled bedroom level from a 95 dBA alarm calculated at about 46.5–76 dBA depending on path
- Distance / position
- Pillow and bedroom positions, modelled from the source
- Frequency weighting
- A
- Conditions
- Varying ambient noise, door position, construction and alarm interconnection
- Supports on this page
- The adult arousal thresholds and the structural explanation of the gap
- Limitations
- The worked attenuation examples are calculated rather than universally field-verified, and the tone-threshold figures are secondary citations.
Sleeping children and smoke alarms, Asia-Oceania Symposium on Fire Science and Technology 4
Bruck, D. and Bliss, A. — Victoria University · 2000 · Peer-reviewed conference paper
- What was measured
- Whether children wake to a standard hallway smoke alarm, with the level received at the child's sleeping position
- Reported level
- 85% of children did not reliably awaken to a standard hallway alarm received at 60 dBA
- Distance / position
- At the child's sleeping position, alarm located in the hallway
- Frequency weighting
- A
- Conditions
- Children in natural sleep, standard high-pitch tone alarm
- Supports on this page
- The child arousal row of the second table and the central finding of the page
- Limitations
- Small experimental sample and a single alarm type and tone.
Developmental auditory arousal-threshold studies
Busby, K. and Pivik, R. T. — reported via CPSC-ES-0502 · 1983 and 1994 · Peer-reviewed research reached through a secondary citation
- What was measured
- Auditory arousal thresholds by age using test tones
- Reported level
- No participant under the age of 15 was woken by tones below an 80 dB average intensity across trials; the oldest subjects were woken at an average of 50 dB
- Distance / position
- At the sleeper in a laboratory setting
- Frequency weighting
- Test-tone dB, not stated as A-weighted
- Conditions
- Age-graded comparison
- Supports on this page
- The age rows of the arousal table
- Limitations
- Read via the CPSC review rather than the original papers, and laboratory tones rather than alarms.
NFPA 72 section 18.4.6, sleeping area audibility requirements
NFPA, as referenced in adopted municipal fire code · 2025 edition referenced · Model code, reached only through secondary municipal references
- What was measured
- Audibility performance required in sleeping rooms, based on ambient level plus a margin at the pillow
- Reported level
- Not quoted: we could not read the operative text, so no numeric formula from NFPA 72 appears on this page
- Distance / position
- Pillow / sleeping area
- Frequency weighting
- Not confirmed
- Conditions
- Systems intended to wake sleeping occupants
- Supports on this page
- The named gap about code audibility requirements
- Limitations
- Listed for transparency about what we could not verify.
Levels on this page are published measurements, quoted with the position and conditions their source states. See our editorial standards for how we source and review noise data.
Test Your Environment
Use Snap Decibel Meter to measure sound levels around you — free, private, and instant.
Start Noise Measurement