SRS airbag warning light: 130,095 cars, lamp did not light

Last updated · OBDCode UK editorial

SRS 130,095 test records → 11.4% (95% CI 11.2–11.5). Compiled by OBDCode UK, edition 2026.1. These are not official statistics and not a DVSA or DVLA statistic. Contains public sector information licensed under the Open Government Licence v3.0.

The SRS / airbag warning lamp is the certificate line “SRS Malfunction Indicator Lamp warning lamp does not illuminate”. Colour on the dash does not say what failed.

If the airbag lamp is on

Limited — drive directly to a garage, no extra journeys. That is not Stop. The lamp can be red and still Limited. Colour does not set drive advice.

What the 2024 fail cohort did next

130,095Cars that failed on this lamp in 2024
11.4% (95% CI 11.2–11.5)Same wording failed again in 2025
34.6% (95% CI 34.3–34.8)Failed 2025 MOT for any reason
14.0 pointsNot tested again vs control (point)

The same-item rate is cars that failed this wording again in 2025, among every car that failed it at the first 2024 normal test. Cars not tested in 2025 stay in the denominator and count as not repeating. The extracts do not record the repair.

The not-tested gap is 14.0 percentage points more than cars that did not fail on this wording. It is a point, not an interval. A missing later test is not a scrappage flag.

Odometer change to the next annual MOT

Among cars with a valid MOT-to-MOT pair, the stored percentiles of odometer change from the 2024 index test to the next annual MOT. n is the pair count, not the fail cohort. These are points. They are not recomputed here, and they are not “miles driven with the light on”.

Published where the treated pair count is at least 500.

Scroll horizontally to read every mileage percentile.

2024 index test Pairs 10th 25th Median 75th 90th
Failed on this line 98,529 1,201 2,981 5,660 9,125 13,332
Did not fail on this line 28,432,344 1,504 3,153 5,745 9,188 13,365

The median on this line is 85 miles below the control median (point).

Days until a later retest passed

Of the 102,653 cars that failed on this line and then passed a later 2024 retest, half did so within 2 days (point). That planning figure excludes same-day station rectification. The retest window count is 130,111 — it is not the headline fail cohort.

Among the 103,058 cars that came back for a later 2024 retest, 99.6% (95% CI 99.5–99.6) passed on the first completed attempt.

A passed retest is a passed retest. The extracts do not say the car was repaired. Day counts are published only when at least 300 cars passed a later 2024 retest.

By fuel

Each row is that fuel only. Electric appears here when the cell reaches the floor. Area PE on the regional table is Peterborough, not petrol.

A rate below 300 cars in that fuel cell prints as —.

Scroll horizontally to read every fuel column.

FuelCarsSame-item failNot-tested gap
Petrol 63,706 11.2% (95% CI 11.0–11.5) +15.0 pp
Diesel 65,074 11.6% (95% CI 11.4–11.8) +12.5 pp
Hybrid Electric (Clean) 523 6.9% (95% CI 5.0–9.4) +4.9 pp
Electric 417 3.1% (95% CI 1.8–5.3) +1.6 pp

By age at the 2024 test

The table is crude — age is still in the number. The age-standardised rate beside it is a weighted sum of five age-specific rates on the Class 4 earliest-2024-NT known-age mix. That standardised figure is not a binomial and does not carry a Wilson interval. A band below 300 cars (crude) or 300 cars (standard) prints as —.

Age-standardised same-item fail: 9.7% (labelled age-standardised). Crude same-item fail: 11.4% (95% CI 11.2–11.5).

Scroll horizontally to read every age column.

AgeCarsCrude same-itemAge-specific rate
3–5 years 3,461 3.2% (95% CI 2.7–3.9) 3.2% age-specific
5–8 years 10,875 8.2% (95% CI 7.7–8.7) 8.2% age-specific
8–12 years 27,500 10.0% (95% CI 9.6–10.3) 10.0% age-specific
12–16 years 33,833 12.4% (95% CI 12.1–12.8) 12.4% age-specific
16 years and over 54,426 12.5% (95% CI 12.3–12.8) 12.5% age-specific

What else was on the same certificate

Other L0 wordings on certificates that already carry this lamp line. Rate = certificates with both ÷ certificates with the lamp. Cells below 500 certificates are omitted, not shown as zero.

130,094 certificates carried this lamp wording.

Scroll horizontally to read every co-occurrence column.

#Also on the certificateCertificatesShare
1 Position lamp A lamp missing, inoperative or in the case of a multiple light source more than 1/2 not functioning 15,239 11.7% (95% CI 11.5–11.9)
2 Headlamp aim The aim of a headlamp is not within limits laid down in the requirements 14,467 11.1% (95% CI 11.0–11.3)
3 Windscreen washers not working or not providing sufficient fluid to clear the windscreen 12,836 9.9% (95% CI 9.7–10.0)
4 Malfunction indicator lamp Engine MIL illuminated indicating a malfunction 11,800 9.1% (95% CI 8.9–9.2)
5 Tread depth worn close to legal limit/worn on edge 11,423 8.8% (95% CI 8.6–8.9)
6 Ball joint has slight play 11,088 8.5% (95% CI 8.4–8.7)
7 Joints A transmission shaft constant velocity joint boot missing or no longer prevents the ingress of dirt etc 9,300 7.1% (95% CI 7.0–7.3)
8 Pins and bushes pin or bush worn but not resulting in excessive movement 9,065 7.0% (95% CI 6.8–7.1)
9 Stop lamp missing, inoperative or in the case of a multiple light source more than 1/2 not functioning 8,613 6.6% (95% CI 6.5–6.8)
10 Wipers Wiper blade missing or obviously not clearing the windscreen 8,259 6.3% (95% CI 6.2–6.5)

Models with the highest same-item fail rate

Among models with at least 200 cars in this fail cohort. Age-adjusted recurrence is used where that field exists on the row; otherwise the crude fail_same_rate is printed and labelled crude. Below the floor the rate is —.

Scroll horizontally to read every model column.

#Make and modelCarsSame-item fail
1 SEAT Mii 202 22.8% (95% CI 17.5–29.0) crude
2 Nissan Almera 438 21.0% (95% CI 17.5–25.1) crude
3 Mitsubishi Colt 378 17.7% (95% CI 14.2–21.9) crude
4 Volkswagen Passat 2,046 16.7% (95% CI 15.2–18.4) crude
5 Renault Kangoo 1,012 16.6% (95% CI 14.4–19.0) crude
6 BMW 120 297 16.5% (95% CI 12.7–21.1) crude
7 SEAT Altea 287 16.4% (95% CI 12.5–21.1) crude
8 Vauxhall Vivaro 2,016 16.3% (95% CI 14.7–17.9) crude
9 Nissan Qashqai 6,522 16.1% (95% CI 15.2–17.0) crude
10 Renault Trafic 973 16.0% (95% CI 13.9–18.5) crude
11 BMW 1 Series 200 16.0% (95% CI 11.6–21.7) crude
12 Nissan Micra 1,982 16.0% (95% CI 14.4–17.7) crude

Largest postcode areas in this fail cohort

Postcode area of the 2024 index test. Areas below 300 cars are omitted. Area PE is Peterborough, not petrol.

Scroll horizontally to read every postcode-area column.

Postcode areaCarsSame-item fail
CF 3,134 13.5% (95% CI 12.3–14.7)
BS 3,061 14.0% (95% CI 12.8–15.3)
B 3,056 10.9% (95% CI 9.9–12.1)
LE 2,620 6.1% (95% CI 5.3–7.1)
NG 2,483 10.6% (95% CI 9.5–11.9)
PO 2,435 13.4% (95% CI 12.1–14.8)
S 2,339 11.0% (95% CI 9.8–12.4)
EX 2,200 14.0% (95% CI 12.7–15.6)

Two-repeat and never-seen-again

Of 135,060 cars that failed this wording on their earliest 2023 annual test, 11.9% (95% CI 11.8–12.1) also failed on the same wording in 2024, and 2.5% (95% CI 2.4–2.6) failed on it in 2023, 2024 and 2025. This is consecutive years of the same certificate line, not a garage outcome.

Of 135,060 cars that failed this wording on their earliest 2023 annual test, 66.5% (95% CI 66.2–66.7) appear in the 2025 Class 4 extract at all (33.5% (95% CI 33.3–33.8) do not). The complement of seen_2025_rate is not_seen_2025_rate, not a gone rate. A missing later test is not scrappage.

Published where the 2023 origin is at least 500 cars.

How these figures are calculated

Window: first normal 2024 Class 4 test (test type NT) → 2025 outcome. The cohort is every vehicle whose index test failed this wording. The control is every other Class 4 vehicle whose index test did not fail this wording. Counts are test records. vehicle_id stability across annual extracts is an empirical finding, not a DVSA guarantee. Cars not tested in 2025 stay in the denominator and count as not repeating.

Floors stated on this page: cohort 500; model 200; age / fuel / postcode / band 300; age-standard band 300; miles pairs 500; returned 300; passed retest 300; three-year chain 500; co-occurrence 500.

Binomial rates print as a percentage with a 95% Wilson interval, or as —. Lifts, gaps, medians and percentiles stay points. Full method and the Replication section: how these figures are calculated.

Download the tables for edition 2026.1 from the dataset page.

These figures are a statistical association from the national MOT extracts. They are not a prediction about this car, and we have not inspected it.

Source: DVSA Anonymised MOT tests and results, 2024 and 2025 extracts (Class 4). Repeat rate = cars that failed on this wording again in 2025 ÷ unique cars that failed on it at their first 2024 normal test. Contains public sector information licensed under the Open Government Licence v3.0.