Road risk in Swindon
Swindon ranks 182 of 192 British highway authorities on our road risk score. Here is what the official record shows, and how it compares.
Police recorded 401 injury collisions in Swindon in 2025, hurting 511 people, of whom 65 were killed or seriously injured and 1 were killed. Set against the traffic its roads actually carry, that is 58.9 killed or seriously injured per billion vehicle miles, 27% below the median ranked authority (80.5).
- Collisions
- 401 in 2025
- People hurt
- 511 casualties
- Killed or seriously injured
- 65
- Killed
- 1
- KSI per billion vehicle miles
- 58.9 (median 80.5)
- KSI, three-year average
- 53 a year
- Severity-adjusted KSI, 2025
- 76.8 (recorded: 65)
- Severity-adjusted KSI, three-year average
- 74.8 a year (recorded: 53)
- KSI per 100,000 residents
- 26.5 — rank 155
- Population
- 245,286
- Risk score
- 25.3/100 — rank 182 of 192
Compiled by Stephen Hall, Data lead, Swift Assist. Source: DfT STATS19 and DfT road traffic estimates (2025), OGL v3. Updated 2026-10-08.
Five years of collisions in Swindon
Across 2021–2025, recorded collisions were broadly flat, with 2022 the worst year of the period at 425 collisions.
Collisions per year in Swindon
2025: 401 (up 5% on the year before) · peak 425 in 2022 · -2% across the period
Axis clipped to the data range to show year-on-year movement.
Swindon looks less dangerous per mile driven than per resident
Divide casualties by the people who live here and Swindon ranks 155. Divide by the miles actually driven on its roads and it ranks 182 — a difference of 27 places.
That gap runs the other way: relative to the traffic its roads carry, Swindon does less harm than its casualty count per resident implies. Places with heavy local travel and few through routes tend to look worse per head than the volume of driving warrants. Neither number is the true one. Which you should use depends on whether you are asking how safe it is to live somewhere or how safe it is to drive there.
Vehicle crime · year ending June 2026
Vehicle crime in Swindon
3.11
offences per 1,000 residents
758
offences recorded, 12 months
171
of 287 areas, worst first
Swindon recorded 3.11 vehicle offences per 1,000 residents, below the England and Wales rate of 5.04.
Rate calculated from 758 offences and a population of 243,900 (ONS mid-2024), both taken from the same ONS release. Contains public sector information licensed under the Open Government Licence v3.0. Source: Office for National Statistics, Crime in England and Wales: Police Force Area data tables, year ending June 2026; police recorded crime collected by the Home Office.
Pedestrians, cyclists and riders in Swindon
These are the road users with the least protection when something goes wrong, and they carry a disproportionate share of the serious injuries. Across 401 collisions in Swindon, the three groups appear 160 times between them — an involvement rate of 39.9%.
| Road user | Collisions | Share of all collisions |
|---|---|---|
| Pedestrians | 42 | 10.5% |
| Cyclists | 69 | 17.2% |
| Motorcyclists | 49 | 12.2% |
A single collision can involve more than one of these groups and is then counted in each row, so the rows do not add up to a share of the total.
When collisions happen in Swindon
DfT records the light level and road surface at every collision, but publishes neither below national level. These are Swindon’s own shares over 1,188 collisions, with the national figure beside each one for comparison.
| Share of collisions | Swindon | National | How it reads |
|---|---|---|---|
| In darknessAny collision where the officer recorded darkness, lit or unlit. | 29% | 28.7% | close to the national figure |
| In darkness, no street lightingThe rural lighting picture rather than the urban speeding one. | 6% | 6.3% | close to the national figure |
| On a wet road surfaceWet, damp or standing water at the scene. | 26.2% | 24.5% | close to the national figure |
| On snow or iceRare everywhere, and concentrated in a handful of authorities. | 1.4% | 1.4% | close to the national figure |
On the darkness share, Swindon ranks 71 of 192 authorities, counting from the highest.
The winter effect. Collisions in Swindon run 10.5% higher from November to January than from May to July — against 4.5% lower nationally. That is 325 collisions against 294.
What this does not tell you
- A share is not a rate. A high darkness share means darkness accounts for more of Swindon’s collisions — not that its roads are more dangerous overall. Read it alongside the casualties-per-mile-driven figure above.
- Condition is recorded, not tested. The officer records the surface and light level at the scene. It is an observation, not a measurement, and it is not a finding that the condition caused the collision.
- Exposure is not controlled for. Fewer people travel at 3am than at 3pm, so a collision in darkness is drawn from a smaller pool of journeys. Without hourly traffic counts, these shares describe when collisions happen rather than how risky an hour is.
Every authority compared on the same measures: winter road risk across Britain.
Swindon against similar authorities
A national rank compares Swindon with places that may be nothing like it. This is the comparison a council actually uses: its statistical nearest neighbours — the authorities ONS identifies as most similar on population, economy and demography, restricted to the same tier, so counties are compared with counties. Within this group Swindon is 4 of 8 on casualties per mile driven, against a group median of 58.5.
| Authority | KSI / bn miles | Score | National rank |
|---|---|---|---|
| Medway | 137.6 | 58.5 | 65 |
| Bexley | 123 | 53 | 84 |
| Bracknell Forest | 67.5 | 44.7 | 124 |
| Swindon (this page) | 58.9 | 25.3 | 182 |
| Cheshire West and Chester | 58.1 | 38 | 149 |
| South Gloucestershire | 42.4 | 35.8 | 155 |
| West Northamptonshire | 37.5 | 33.3 | 164 |
| West Berkshire | 33.9 | 34.6 | 158 |
Peer group: ONS statistical nearest neighbours, Open Government Licence v3. ONS notes the grouping is a snapshot that is not comparable with earlier editions, and is not a judgement about any authority’s performance. The road figures against it are ours.
Swindon against its closest comparators
Swindon sits 182 of 192 nationally, and 142 of 148 within England. These are the authorities nearest to it in the table.
Road risk in Swindon, answered
How dangerous are the roads in Swindon?
In the latest official DfT figures, 401 road collisions in Swindon injured or killed 511 people, 65 of them killed or seriously injured. Measured against the traffic its roads actually carry, that is 58.9 people killed or seriously injured per billion vehicle miles, against a median of 80.5 across the ranked authorities — placing Swindon 182 of 192 authorities on our combined score, where 1 is the highest risk. The score weighs that rate alongside who is being hurt, the five-year trend and how often a serious collision proves fatal, so the rank is not the rate alone.
Are road casualties in Swindon rising or falling?
Across 2021–2025, recorded collisions in Swindon were broadly flat, with 2022 the worst year of the period at 425 collisions. Year-to-year movement in a single authority is noisy, so the five-year direction carries more than any single year's change.
Who is being hurt on Swindon's roads?
In Swindon there were 42 collisions involving pedestrians, 69 involving cyclists and 49 involving motorcyclists, out of 401 in total. Added together that is an involvement rate of 39.9%, though a collision involving two of those groups is counted in both figures, so it is a comparable index rather than a clean share.
Who is responsible for road safety in Swindon?
Swindon is a unitary authority, and is the highway authority for its own roads. That means it sets local speed limits, designs junctions and crossings, and decides safety scheme spending. Motorways and most trunk roads are managed separately by National Highways in England, Transport Scotland, or the Welsh Government, so the worst road in an area is often not the council's to change. Enforcement sits with the local police force.
Where does this data come from?
DfT STATS19, the record police complete at the scene of every injury collision, published annually under the Open Government Licence, combined with DfT's own road traffic estimates for the mileage denominator. Collision, casualty and KSI counts on this page are for 2025; only the trend compares 2021–2025. STATS19 counts collisions police attended and recorded, so collisions never reported do not appear, and under-reporting is known to be worst for cyclists and for slight injuries.
What actually changes the odds in Swindon
Data is only useful if it changes what you do. These three are picked from Swindon’s own figures above, each with the number that prompted it. Every guide explains how the thing works before it suggests buying anything, and most cost less than a third of a typical insurance excess.
Helmet camera
£50–£200
17% of collisions in Swindon involve someone on a bike. In a disputed pull-out, footage from the rider's viewpoint is what settles liability.
How it works →
Alarmed disc lock
£20–£50
Motorcyclists are in 12% of Swindon's collisions, and bikes are also lifted rather than ridden away. An alarmed disc lock covers the theft half for about £30.
How it works →
We are a data business, not a retailer: these link to our guides, which link on to retailers. Some of those links earn us a small commission at no cost to you, and we hold no stock and take no payment for placement. All security guides.
More on Swindon and UK road risk
Where collisions cluster in Swindon
The interactive hotspot map, down to individual junctions.
Read more →All 192 British authorities ranked
The full sortable league table, with the method behind it.
Read more →Blackspots that were never fixed
Sites with injury collisions in every one of the last five years.
Read more →UK road risk: the national picture
What the official record says about Britain's roads overall.
Read more →Need this cut differently?
Custom analysis by ward, corridor or junction. Free for campaigns and press.
Read more →Free API
Pull these figures programmatically, with provenance on every endpoint.
Read more →