Great Britain Road Casualties 2025: What July’s Final Figures Show

What The Final Figures Say

What changed in Great Britain's 2025 totals?

The final release records 1,538 reported fatalities, an adjusted national KSI series of 29,918 based on police-reported data, and 127,883 reported casualties of all severities. Compared with 2024, fatalities fell 4%, KSI rose 4%, and all-severity casualties changed little.

STATS19 is the system through which police forces supply information about reported personal-injury road collisions to the Department for Transport. Its geographic scope here is Great Britain: England, Scotland and Wales. Northern Ireland is not included. The three headline measures overlap rather than add together. KSI means killed or seriously injured, while the all-severity total also includes slight injuries. The published KSI figure is an adjusted national series designed to account for changes in injury-severity reporting by some police forces; it should not be described as a simple raw police count. “Final” identifies the status of the main annual release, not universal coverage of every road event.

That separation explains why the latest annual picture can be mixed without being contradictory.

That separation explains why the latest annual picture can be mixed without being contradictory.

Why The Picture Is Mixed

Can fatalities fall while serious casualties rise?

Yes. Fatalities and KSI casualties are distinct severity measures, so they can move in different directions during the same year. The fall in deaths and rise in adjusted KSI are both valid annual findings. Neither should be used alone as a complete verdict on road safety.

Time period changes the interpretation too. Compared with 2004, fatalities have more than halved, KSI casualties have fallen by around one third, and all-severity casualties have fallen by over half. Those longer trends do not erase the 2025 KSI rise or prove that progress was even across every year, place or road-user group. Exposure adds a third lens: the final release reports 4 fatalities, 86 KSI casualties and 369 casualties of all severities per billion vehicle miles in 2025. These national rates relate harm to overall motor-vehicle travel. They are not the probability attached to a particular driver, road or trip.

Three-lens check:

1. Latest year: identify which severity measure moved.
2. Long term: compare like definitions over the stated period.
3. Exposure: distinguish a per-mile rate from an individual forecast.

This method is more useful than selecting whichever number appears most reassuring or alarming.

This method is more useful than selecting whichever number appears most reassuring or alarming.

Who Faces The Highest Exposure

Which road users stand out in 2025?

Car occupants formed 53% of all reported casualties. Pedestrians, motorcyclists and pedal cyclists together formed over half of KSI casualties. Motorcyclists accounted for 25% of road deaths but 1% of road traffic and had the highest death and serious-injury risk per mile.

Those percentages use different bases. The 53% figure is the share of casualties of all severities who were car occupants. “Over half” concerns three road-user groups combined within the narrower KSI population. The motorcycle comparison places a share of deaths beside a share of traffic, then adds an exposure-adjusted per-mile finding. Putting all these percentages into one ranking would mix incompatible denominators. The practical lesson is proportionate: raw totals should be checked against severity and exposure before they inform a discussion about training, routes or vehicle use. The national evidence does not predict an individual rider's outcome, explain why a particular collision happened or establish legal responsibility.

Denominator guide: all casualties, KSI casualties, deaths, traffic share and rate per mile answer different questions. Keep the label attached whenever a figure is repeated.

Denominator guide: all casualties, KSI casualties, deaths, traffic share and rate per mile answer different questions. Keep the label attached whenever a figure is repeated.

Where Severe Collisions Concentrate

Which road environments need the closest attention?

Rural roads accounted for 60% of reported fatalities and 45% of traffic in 2025. Motorways accounted for 6% of fatalities and 21% of traffic. Single carriageways accounted for 77% of fatalities, 78% of KSI casualties and 73% of all casualties.

These are national distributions, not danger labels for every road. The rural-road and motorway comparison places fatality share beside traffic share, which helps show where fatal outcomes were concentrated relative to travel. The single-carriageway percentages describe shares across three severity measures. Categories can overlap: a single carriageway may also be rural or urban. The figures therefore should not be added as though they describe separate events. Nor do they isolate the effect of road design, speed, weather, traffic mix or behaviour. A proportionate practitioner response is to treat the national pattern as a prompt for reviewing relevant local or route-specific evidence, not as a substitute for it.

Environment check: use “accounted for” or “was recorded on”; avoid “was caused by”. Description can guide closer attention, but causation needs evidence about the actual event.

Environment check: use “accounted for” or “was recorded on”; avoid “was caused by”. Description can guide closer attention, but causation needs evidence about the actual event.

What The Data Cannot Prove

What should readers avoid concluding from STATS19?

STATS19 does not capture every collision or casualty because not every personal-injury collision must be reported to police. Recorded road-safety factors apply to a subset, may overlap and do not establish sole cause or legal fault. Planned September outputs were not complete within this source boundary.

Coverage is the first limitation. The published totals describe police-reported personal-injury collisions within the STATS19 scope, not a census of all road harm. The retained evidence acknowledges under-reporting but does not quantify the missing share, so no percentage should be estimated. The “fatal four” statistic has a separate boundary: almost two thirds of fatal collisions with recorded road-safety-factor data had at least one such factor assigned. Factors are officer-recorded for a subset and more than one may apply to the same collision. They can describe recurring circumstances, but they do not prove sole causation, civil liability or criminal responsibility. Finally, July is the main annual final release; September analysis and product work were planned, not evidenced here as complete.

Shows / misses / cannot prove: STATS19 shows reported patterns; misses unreported events; and cannot decide an individual's future risk, legal fault or vehicle-damage outcome.

Shows / misses / cannot prove: STATS19 shows reported patterns; misses unreported events; and cannot decide an individual's future risk, legal fault or vehicle-damage outcome.

What Happens Next

How does the 2026 strategy change context?

The 2026 Road Safety Strategy sets a Great Britain goal horizon for reducing killed or seriously injured casualties by 2035. Its four themes cover road users; technology, innovation and data; infrastructure; and enforcement. These are future policy goals, not achieved outcomes or guarantees.

The statistics and strategy perform different jobs. The final 2025 figures provide a baseline and describe reported outcomes before much of the strategy's horizon. The strategy organises intended action and future monitoring. Its existence cannot retrospectively explain the 2025 movements, prove that a particular intervention works or guarantee a reduction by 2035. The four themes also support a multi-factor reading of road safety: action can involve road users, technology and data, infrastructure and enforcement, rather than one universal cause or solution. Future assessment needs later comparable figures, transparent definitions and disclosure of methodological changes.

Practical monitoring rule: preserve the measure, geography, baseline and denominator. When later data arrives, compare like with like before declaring progress or decline.

The responsible conclusion is measured: long-term improvement, a mixed latest year, uneven exposure, clear data limits and a policy horizon whose outcomes remain to be demonstrated.

Use these figures as a disciplined reading framework rather than a verdict on one journey. Keep the severity measure, comparison year, geographic scope and denominator beside every number. When new analysis appears, check whether definitions or methods changed before comparing it with 2025. For an actual collision, use evidence from that event and the appropriate official or professional process; national statistics remain context, not a decision about cause, liability, eligibility or outcome.

Use these figures as a disciplined reading framework rather than a verdict on one journey. Keep the severity measure, comparison year, geographic scope and denominator beside every number. When new analysis appears, check whether definitions or methods changed before comparing it with 2025. For an actual collision, use evidence from that event and the appropriate official or professional process; national statistics remain context, not a decision about cause, liability, eligibility or outcome.

FAQs