How Loud Is a Car Horn? What Type Approval Actually Requires

UN Regulation 28 governs audible warning devices in the many countries that apply UNECE vehicle regulations. It separates the device from the installation. The device is measured in a controlled setup at 2 metres, which is how a horn is approved as a component. The complete vehicle is then measured at 7 metres, which is how the installed system is checked.

Because level falls with distance, and because mounting position, bodywork and shielding all matter, the same horn produces a different figure in each test. Quoting one as "how loud a car horn is" without saying which test it came from is where most published numbers go wrong.

NHTSA explicitly states it does not require a horn on motor vehicles; FMVSS 101 addresses the control if a horn is provided. This is not a statement about state road-use rules. Federal commercial-vehicle provisions separately require horns for the vehicle categories covered by 49 CFR 393.81. We found no federal numeric horn-output requirement for passenger cars in the cited NHTSA material.

Do not generalise the federal passenger-car finding to every US jurisdiction or vehicle category. Voluntary industry practices and local requirements are separate from UNECE type approval.

Historical horn measurements are not used here as a numeric reference: the original study, weighting and setup have not been established in this page's evidence ledger.

This evidence gap should not be filled with a generic estimate for modern cars.

None of the documented figures answers the pedestrian's question. Your position relative to the horn's axis matters, as does reflection from buildings, the distance, and whether the horn is a single tone or a pair.

Horns are also brief. Level alone is a poor description of a short, startling event, which is part of why exposure frameworks treat impulsive sound separately from continuous noise.

The design intent explains the acoustics. A horn has to be recognised through traffic noise, through a closed vehicle, and against whatever a driver is listening to, which is why horns use paired tones and a spectrum concentrated where hearing is sensitive rather than simply maximum level.

That also explains why a horn heard from behind can seem quieter than one heard head-on at the same distance: the device is directional, and approval measurement fixes the position precisely for that reason.

Where horn use is a nuisance — a junction, a taxi rank, a level crossing — the regulation that applies is usually a local traffic or nuisance rule about when a horn may be sounded, not a decibel limit on the device.

That is a sensible division. Approval already controls what the device can emit; what remains is behaviour, and behaviour is regulated by time-of-day and permitted-use rules rather than by measurement at a complainant's window.

If you want to document a horn problem, the useful record is frequency and timing of events over days, not a single peak reading, because the disturbance is caused by repetition rather than by one level.