How Loud Is a Police Siren? Decibel Levels Explained

Warning sirens exist to be heard over traffic inside a vehicle with the windows up and a radio on. That makes them one of the few sounds engineered to a required loudness, tested on a bench rather than in the field.

The test geometry is the key to reading any siren figure: microphone at 3 metres, on the loudspeaker's axis, with the requirement stepped down in 10-degree increments to 50 degrees off axis. Both a minimum and a maximum apply — the minimum so the warning works, the maximum so it does not needlessly injure hearing.

The same nominal function is governed by different bodies with different ceilings, and the differences are documented rather than mysterious. The NIJ guide attributes them to different loudspeaker configurations — single versus dual — and different averaging times in the measurement, such as a 0.02-second versus 0.125-second detector setting.

That matters if you are comparing devices across markets. A 123 dB(A) figure from an ambulance specification and a 118 dB(A) figure from an industry recommended practice are not describing different physics; they are describing different rules.

A siren aimed forward still puts substantial energy into the vehicle cabin, and the people most exposed to it are the crew, repeatedly, for a career. SAE J1849 addresses this directly by recommending that in-cabin level at the riding positions not exceed 85 dB(A).

That 85 dB(A) recommendation is not a coincidence — it is the level occupational guidance uses as an 8-hour reference. Placing it in a siren standard is an explicit acknowledgement that the warning device is also an occupational exposure.

A peer-reviewed study from VSB–Technical University of Ostrava, examining warning devices used by the Czech police, reports a national legal minimum of 105 dB(A) under Czech Decree 341/2002 and a maximum of 118 dB(A) under UN/ECE Regulation 28.

The upper figure matches SAE J1849's ceiling exactly, which is reassuring for anyone wondering whether these numbers are arbitrary. The same study states that officers operating sirens experience equivalent levels above 105 dB, but gives no measurement distance, duration or method for that claim, so we are treating it as context rather than evidence.

This is the weakest part of the record. Occupational dosimetry studies of police and ambulance crews exist — a NIOSH report on highway patrol officers and a 2015 study of patrol officer daily noise exposure in Wisconsin — but we could not extract their numeric results, so we are not quoting figures from either.

We also found no source giving the level of a siren at a realistic street distance, say 50 or 100 metres. Every verified number here is a 3-metre bench measurement, and we are not going to propagate it outward with a formula: a siren mounted on a moving vehicle in a street with buildings does not behave like a point source in free space.

A passing siren is a rapidly changing, directional, tonally sweeping sound. It is genuinely interesting to watch on a meter and genuinely unsuited to a single number.

Our browser meter reports A-weighted digital signal level in dBFS when uncalibrated — a relative figure, not sound pressure — and an estimated dBA figure after a one-time local calibration. Either way it cannot verify a bench specification measured at 3 metres on axis at a controlled supply voltage, and a phone microphone close to a siren is likely to be compressing.