How Loud Is a Stadium Crowd? What the Measurements Actually Show

A stadium study of a vuvuzela-heavy match environment recorded an average level of 100.5 dB(A) with a peak of 140.4 dB(C). Those two numbers describe the same environment in two ways and both matter.

The 100.5 dB(A) average is the hearing-relevant one, because it is sustained. A level in that region maintained over a match is a substantial exposure by any occupational standard. The 140.4 dB(C) peak is impulsive — a momentary pressure spike, C-weighted and peak-detected, which is a fundamentally different measurement from the average and cannot be compared with it.

This is one context in one study, and the vuvuzela is a specific and unusually consistent noise source. We are not presenting it as the level of stadiums in general.

The 142.2 dB(A) stadium record is one of the most repeated figures in noise trivia, and it answers a different question than the one people ask it.

A record attempt is optimised: the crowd is prompted, the moment is coordinated, the microphone placement is chosen and the metric is a momentary maximum. The methodology of this particular attempt has also been disputed. None of that makes it fraudulent — it makes it a record, which is a competitive achievement rather than a characterisation of a normal match.

If you want to know what a stadium is like for the people in it for two hours, the sustained average is the number to look at, and it is roughly 40 dB lower.

Two structural factors combine. Tens of thousands of people are each a source, and crowd noise coordinates itself — chants, reactions to play, and the simple fact that people raise their voices when everyone else does.

Then there is the bowl. A stadium with a roof or steep enclosed stands reflects sound back onto the crowd instead of letting it escape, and the crowd responds to its own amplified noise. Open-bowl stadiums are consistently reported as quieter than enclosed ones for the same attendance, which is a design effect rather than a difference in enthusiasm.

NIOSH recommends an occupational limit of 85 dBA as an 8-hour time-weighted average with a 3 dB exchange rate, meaning each 3 dB increase halves the time considered acceptable. That is an occupational criterion and we present it as educational context, not as a personal exposure calculator.

Read against a sustained 100.5 dB(A), the direction is unmistakable without arithmetic: this is a level where hearing protection is a reasonable precaution, particularly for people who attend regularly, and for stewards and staff who are there as a job. The impulsive 140.4 dB(C) peak is a separate concern, since very high peaks can damage hearing without any duration at all.

We have one measured context, not a survey. We found no dataset comparing sustained levels across stadium designs, attendances or sports with a consistent method — which is exactly what would be needed to say how loud a typical stadium is.

Nor did we find seat-position data, so we cannot tell you whether the upper tier is louder or quieter than the touchline.

A phone is a reasonable tool for the sustained level and a poor one for the peaks.

Uncalibrated our meter reports A-weighted digital signal level in dBFS, which is relative rather than sound pressure; after a one-time local calibration it reports an estimated dBA figure. At crowd-roar levels a phone microphone is at the top of its range and will compress, so peaks in particular will read low. Our meter also does not report a C-weighted peak, so a figure like 140.4 dB(C) is outside what it can produce.