Health effects
Sleep, blood pressure, hearts and children's attention: what the WHO evidence rates, and how strongly.
See the evidence
The sound itself
A car horn is allowed to measure up to 112 dB(A) seven metres in front of the car. In a queue at a red light, everyone around it hears it. This page explains what that does, in numbers you can check.
India's horn standard, AIS-014, allows a car horn to measure 93 to 112 dB(A) at seven metres in front of the vehicle; horns on small two- and three-wheelers may measure 83 to 112 dB(A). A horn that loud is legal.
For comparison, the American Academy of Audiology lists a car horn at 110 dB, in the same row as a concert. Close to the vehicle, say around 100 to 110 dB. An IIT Madras team measuring harsh honks at six Chennai sites found them averaging 116 dB(A) at two metres.

The scale is logarithmic. Every 3 dB increase doubles the sound energy; every 10 dB increase is ten times the sound energy and is heard as roughly twice as loud. So a 100 dB horn is not a bit louder than a 70 dB conversation; it carries a thousand times the energy.
Decibels with A-weighting, a filter that follows the ear's sensitivity. Indian standards, WHO and NIOSH all use A-weighted values.
The equivalent continuous A-weighted level over a stated period: the steady level with the same sound energy as the varying noise actually measured. Indian limits are Leq for day (6 am to 10 pm) and night (10 pm to 6 am).
| Source of sound | Level, dB(A) |
|---|---|
| Whisper | 30 |
| Quiet library | 40 |
| Normal conversation | 60 to 70 |
| Motorcycles | 80 to 110 |
| NIOSH limit, 8 hours a day | 85 |
| Concerts, sports events | 94 to 110 |
| Legal car horn at 7 m | 93 to 112 |
| Car horn, close to the car | 100 to 110 |
| Sirens | 110 to 129 |
| Fireworks | 140 to 160 |
NIOSH, the US occupational health institute, recommends no more than 85 dBA averaged over an eight-hour day, with a 3 dB exchange rate. From those two numbers: 88 dBA is safe for four hours, 94 dBA for one hour, 100 dBA for 15 minutes, 112 dBA for under a minute.
A rule of thumb: if you must raise your voice to be heard at arm's length, the noise is probably at or above 85 dBA; if you must shout, 95 dBA or more.
| Level, dBA | Safe time a day |
|---|---|
| 88 | 4 h |
| 91 | 2 h |
| 94 | 1 h |
| 97 | 30 min |
| 100 | 15 min |
| 103 | about 7.5 min |
| 106 | under 4 min |
| 109 | under 2 min |
| 112 | under 1 min |
In Mumbai, IIT Bombay researchers who rode six routes with sound meters found commuters exposed to more than 90 dB daily because of honking, and vehicle occupants above 70 dB for half the commute. In Kolkata, a traffic constable stands in it for a shift.
Hearing studies of traffic police vary by city, year and method: 21 per cent with noise-induced hearing loss in a 1997 Chennai sample of 100; 22 per cent in a 2015 Jammu study of 150. Among 100 Bengaluru auto-rickshaw drivers tested, 40 per cent had noise-induced hearing loss, rising with age and working hours.

Mumbai's Joint Commissioner of Police (Traffic) said in 2023 that motorists wrongly believe honking speeds up traffic. Kolkata's Commissioner said in 2020 that most honking stems from lack of awareness, and that people honk even when traffic is moving.
A 2016 survey for a corporate client in Bengaluru, with 80 observers at 80 junctions, judged 93.7 per cent of horn use to be for no reason; 42 per cent of drivers said they honk because traffic is slow. The common answer: if I don't honk, somebody else will.
A Nagpur study on three roads found honking added 2 to 5 dB(A) to the average level, and that honk counts, not traffic volume, predicted how loud a road was. In Kolkata the Jadavpur University team concluded that avoiding honking could lower levels by about 2 to 5 dB(A). Remember that 3 dB is a doubling of sound energy.
When Kolkata students counted honks outside schools and hospitals with traffic police present, an NGO count found about 60 per cent fewer.

Share this page
Sleep, blood pressure, hearts and children's attention: what the WHO evidence rates, and how strongly.
See the evidence