Official source
Road safety open data
UK Department for Transport
- 2024 rows
- 100,927
- 2025 provisional rows
- 48,472
- Licence
- Open Government Licence v3.0
- Imported
- 2026-07-25
Reported collisions and traffic evidence
UK road context for cycle-trip research
149,399 police-reported personal-injury collision records and 600,551 annual traffic-flow observations are held with their source status intact. This report describes coverage; it does not rank routes as safe or unsafe.
149,399Collision records
22,784Cycle-involved collisions
600,551AADF observations
46,754Traffic count points
24 located GB route records now show fixed-radius final-2024 cycle-involved collision context and the nearest matched 2025 traffic count point. 24 have a traffic match; every page labels the result as representative-point context rather than whole-route conditions.
It means the collision record lists at least one pedal cycle among the vehicles involved. It is not a count of cyclist casualties. These are STATS19 personal-injury collisions reported to police on public roads; unreported incidents are outside the dataset.
| Period | Status | All collisions | Cycle-involved | Fatal severity* | Serious severity* | Slight severity* |
|---|---|---|---|---|---|---|
| 20242024-01-01–2024-12-31 | Final | 100,927 | 15,042 | 85 | 3,768 | 11,189 |
| 20252025-01-01–2025-06-30 | Provisional | 48,472 | 7,742 | 52 | 2,026 | 5,664 |
*Severity is the collision classification, not a casualty count. The 2024 file is the latest final validated full year in this holding. The 2025 file covers January–June and is provisional, unvalidated and subject to revision or duplicate removal.
Traffic exposure evidence
600,551 annual records cover 2000–2025 across 46,754 count-point records. AADF estimates the average vehicles passing a point in 24 hours over a year.
Not live traffic
Not every road link is manually counted every year. The series combines counts and estimates, and point coverage changes, so raw averages are not published here as a national cycling trend.
Not a route score
A collision point near a route does not establish causation or current risk. Meaningful comparison needs exposure, network design, route length, time period and reporting-quality controls.
The held planning-data extract contains 362,520 authoritative NaPTAN access-node rows. Its observed latitude range is 49.90–56.00, so CycleTripPlanner treats it as England and Wales coverage and does not imply that it supplies Scottish Highlands gateways.
The public destination projection keeps nearby rail, coach, ferry and airport records only. Ordinary stop records are not turned into thin pages.
Official methodology
Road traffic estimates
UK Department for Transport
- AADF rows
- 600,551
- Count points
- 46,754
- Licence
- Open Government Licence v3.0
- Refresh
- Source state checked daily
The scheduled job re-counts the live holdings and records their import metadata every day. It does not relabel an old official release as newly published; final and provisional status remains visible until the source really changes.