Food safety statistics
Food Poisoning Statistics UK
Latest complete food poisoning surveillance, historical UK burden estimates and food safety advice, reviewed October 2026. Clear dates, geography and case definitions.
Read more →The road safety picture · 2026 edition
Drink driving
statistics UK
The numbers. The risk.
The journey home.
Explore the latest published UK drink-driving statistics, with casualty, police testing and conviction figures, downloadable charts and original sources.
By Mark McShane
Last checked 16 October 2026 · Latest published evidence
Estimated deaths
220people killed in drink-drive collisions.
Great Britain · 2024 · 95% interval: 190–250All casualties
6,350people killed or injured.
Great Britain · 2024 · All severitiesShare of road deaths
14%involved an over-limit driver or rider.
Great Britain · 2024 · Rounded estimate01
The human cost
Every number represents a person.
These estimates cover reported Great Britain collisions involving at least one driver or rider over the legal alcohol limit. They describe involvement, rather than establish that alcohol caused each collision.
Great Britain · Fatalities · 2024
220 estimated deaths, with uncertainty.
Estimated fatalities · scale from 0 to 300
View chart data
| Measure | People |
|---|---|
| Central estimate | 220 |
| 95% confidence interval | 190–250 |
The wider road safety picture
About one in seven reported road deaths.
14%of road deaths
The share relates to all people killed in these collisions, including passengers and other road users.
The central fatality estimate fell from 260 in 2023 to 220 in 2024. DfT says the change was not statistically significant.
A lower central estimate does not, by itself, establish that the underlying risk has fallen.
Great Britain · 2024
1,740
People killed or seriously injured
This is part of the total of 6,350 casualties, rather than an additional group. Non-fatal injury figures use DfT’s severity adjustments.
Great Britain · 2024
4,360
Reported drink-drive collisions
A collision can injure more than one person. The figures cover reported personal-injury collisions, rather than every journey made after drinking.
Great Britain · Estimated fatalities · 2014–2024
A decade of estimates, with the uncertainty visible.
Dots show central estimates; amber lines show 95% confidence intervals. Vertical scale: 0–350 deaths.
View all annual estimates
| Year | Estimate | 95% interval |
|---|---|---|
| 2014 | 240 | 210–270 |
| 2015 | 200 | 170–220 |
| 2016 | 230 | 200–260 |
| 2017 | 250 | 220–280 |
| 2018 | 240 | 210–270 |
| 2019 | 230 | 200–260 |
| 2020 | 220 | 190–250 |
| 2021 | 260 | 230–290 |
| 2022 | 300 | 270–330 |
| 2023 | 260 | 230–290 |
| 2024 | 220 | 190–250 |
How to interpret the casualty estimates
1.1 Latest Official Collision Data
- 220 deaths was the central 2024 estimate, with a 95% confidence interval of 190–250.
- 1,740 people were killed or seriously injured.
- 6,350 casualties of all severities occurred in around 4,360 collisions.
These are reported Great Britain collisions involving a driver or rider over the legal limit, not proof that alcohol caused each collision.
1.2 Recent Trends
The fatality estimate fell from 260 in 2023, but DfT says the change was not statistically significant. Total casualties changed little.
1.3 Proportion of All Road Deaths
Drink-drive collisions accounted for around 14% of reported road deaths in 2024.
02
Police activity & court outcomes
Testing and convictions tell different stories.
26 comparable forces · England & Wales
Breath-test totals were almost unchanged.
Recorded screening tests · shared scale 0–200,000
2023167,384
2024167,095
050,000100,000150,000200,000
Among the recorded tests · 2024
24,823positive or refused tests15% of all tests in this dataset
A positive screening result or refusal is not a conviction. Tests measure police activity and the people selected for screening, rather than the proportion of all drivers who drink and drive.
Use the same 26-force coverage when comparing 2023 and 2024. The subset may not represent the rest of England and Wales.
England & Wales · Principal offence
Fewer convictions in the latest year.
Driving or attempting to drive over the limit · scale 0–40,000
202329,532
202427,221
202525,973
010,00020,00030,00040,000
2024 to 2025
A 4.6% fall in this conviction series.
There were 25,973 convictions in 2025, compared with 27,221 in 2024. The series counts the specified principal offence.
Failing to provide a specimen and being in charge over the limit are separate offences. Adding them without checking the counting rules can double-count or mix different measures.
Court outcomes depend on detection, charging, case timing and the principal-offence rules. A fall in convictions does not establish a fall in drink-driving prevalence.
Offence definitions and the full enforcement context
2.1 Breath Tests (Police Enforcement)
- For 26 comparable England and Wales police forces, 167,095 breath tests were recorded in 2024, compared with 167,384 in 2023.
- 24,823 tests (15%) were positive or refused in 2024.
- The analysis excludes 17 forces because of data-quality or recording changes. These totals should not be compared with older all-force totals as though coverage were unchanged.
A roadside screening result is not a conviction. Testing also reflects police activity and the circumstances in which officers request a test.
2.2 Drink Driving Convictions (England & Wales)
Driving or attempting to drive over the limit
DfT’s published Ministry of Justice series records 25,973 convictions in 2025, compared with revised figures of 27,221 in 2024 and 29,532 in 2023. The 2024–25 fall was about 4.6%. This series counts the specified principal offence, not every alcohol-related offence.
Failing to provide a specimen
Failing, without reasonable excuse, to provide an evidential specimen is a separate offence. Its outcomes should not be added to an over-the-limit series without checking the offence definition and counting method.
Being in charge of a vehicle over the limit
Being in charge is distinct from driving or attempting to drive. It has its own legal tests and should be identified separately in a statistical comparison.
Failing to provide a specimen while in charge
The alleged circumstances matter when interpreting this offence. Court data also reflect charging and the principal-offence rules, so a change in convictions is not a direct estimate of a change in prevalence.
03
Geography & patterns
Read the boundaries behind the numbers.
GB
Great Britain
The casualty estimates cover England, Scotland and Wales. Northern Ireland is excluded.
E&W
England & Wales
The conviction series covers both nations. Breath-test figures cover only 26 comparable police forces within them.
UK
United Kingdom
The legal-limit comparison includes all four nations. Scotland has a lower limit than the other three.
December · Comparable forces · 2024
17%
Of the year’s recorded tests
December accounted for this share of all tests in the annual dataset, reflecting seasonal police activity.
Among December tests · 2024
10%
Were positive or refused
This percentage uses December tests as its denominator. It is not the share of all drivers who were over the limit.
Targeted campaigns change who is tested. More tests in a month do not automatically mean a higher underlying risk.
Great Britain · 2024 · Three different measures
Who is involved, and when?
80%
Of collision-involved drink-drivers were male
Among motor drivers and riders whose sex was known. The corresponding share in all reported collisions was 69%. These are composition shares, not population risk rates.
85
Drink-drive collisions per billion car miles for under-20s
The mileage-adjusted rate was 15 across all ages. Under-20s had the highest rate, rather than the largest collision count. This comparison covers car drivers.
9%
Of drink-drive collisions occurred from 11pm to midnight
Compared with 1% between 11am and noon. These shares describe the timing of drink-drive collisions, not the risk per journey or the share of all night-time collisions.
04
Current law · Checked 16 October 2026
One UK. Different legal limits.
The limits below are the law at the review date. They cannot be reliably converted into a fixed number of drinks.
England, Wales & Northern Ireland
80 mg / 100 ml blood
35 micrograms per 100 ml of breath
Scotland
50 mg / 100 ml blood
22 micrograms per 100 ml of breath
A limit is not a target
Alcohol affects people differently, and even a small amount can impair driving. Plan not to drive after drinking.
The legal framework, offences and proposed changes
The Road Traffic Act 1988 provides the main Great Britain offences for driving or being in charge while over the prescribed alcohol limit, and failing to provide a required specimen without reasonable excuse. Northern Ireland has separate road-traffic legislation.
UK legal alcohol limits (reviewed October 2026)
- England, Wales and Northern Ireland: 80 mg of alcohol per 100 ml of blood, or 35 micrograms per 100 ml of breath.
- Scotland: 50 mg per 100 ml of blood, or 22 micrograms per 100 ml of breath.
There is no reliable conversion from these limits to a fixed number of drinks. Alcohol affects people differently. Penalties can include disqualification, a fine or imprisonment, depending on the offence and jurisdiction.
The government’s 2026 motoring-offences consultation closed on 11 May. Proposed changes to England and Wales drink-drive limits are proposals; they do not replace the current limits shown above.
05
From evidence to everyday decisions
Plan the journey before the first drink.
01
Arrange the journey home
Plan transport before going out, so getting home does not depend on driving after drinking.
02
Decide who is driving
If someone will drive, plan for them to avoid alcohol. Do not use the legal limit as a drinks allowance.
03
Do not judge by how you feel
Feeling fit to drive is not a reliable check of alcohol level or its effect on reaction times.
What surveys can tell us about behaviour
Self-reported surveys can help explain attitudes, but results depend on the sample, question and recall period. They should not be presented as a count of offences or directly compared with police test results.
THINK! advises that even a small amount of alcohol can impair driving and reaction times. Plan a journey that does not depend on driving after drinking; feeling fit to drive is not a reliable test of alcohol level.
06
The evidence in perspective
Four checks before sharing a statistic.
What is being counted?
Keep people injured, collisions, screening tests and convictions separate. They are not successive stages of a single group.
Where and when?
Keep the geography, data year and reporting coverage attached to the number. A publication from 2026 can contain the latest available 2024 data.
How certain is the estimate?
Give the confidence interval alongside estimated deaths. A lower central figure does not necessarily indicate a statistically significant change.
What does the change prove?
Changes in enforcement or court activity do not, on their own, prove a change in deterrence or in how often people drive after drinking.
For writers, editors & researchers
Use the evidence.
Keep the context.
Download the charts and source-labelled figures. The data pack includes the dates, geographic boundaries and definitions needed to quote them accurately.
Food Hygiene Certificate (2026). Drink Driving Statistics UK: Latest Data & Trends (2026). Evidence checked 16 October 2026. https://www.foodhygienecertificate.co.uk/blog/drink-driving-in-the-uk-statistics
About this guide
Official evidence, explained clearly.
This guide brings together the Department for Transport, Home Office and Ministry of Justice series. It is a synthesis of published evidence, with the measures kept separate and original sources provided below.
The evidence behind this guide
Sources & references
Checked 16 October 2026. Official statistics and current guidance; publication dates and data periods are labelled throughout.
- Department for Transport — Drink-drive collisions 2024 and 2025 conviction series
- Home Office — Roads policing to December 2024
- Ministry of Justice — Criminal justice statistics to December 2025
- GOV.UK — Drink-drive limits
- GOV.UK — Drink-driving penalties
- Road Traffic Act 1988 — Section 5
- Road Traffic Act 1988 — Section 7
- THINK! — Drink a little, risk a lot
- Department for Transport — RAS2001: current fatalities and confidence intervals, 2014–2024 (ODS, July 2026)
- Department for Transport — RAS2022: car drink-driver collisions by age and mileage (ODS, July 2026)
- Department for Transport — Motoring offences consultation (closed 11 May 2026; proposals, not current limits)
How to read this guide
Deaths in drink-drive collisions are estimated, with a 95% confidence interval. Casualty severity is adjusted by DfT and headline counts are rounded. Police breath tests cover a comparable 26-force subset; convictions use a specified principal offence. None is a direct count of every instance of drink driving. Percentage changes calculated from displayed conviction counts are rounded to one decimal place.
