Metric permalink
Rail cancellations (GB)
Part of the Infrastructure domain. Percentage of scheduled trains in Great Britain that were cancelled or part-cancelled, measuring the reliability of the national rail network.
Latest data
Latest reading
2025–26
Change between the two latest observations (2024 to 2025–26), in stress score.
A higher reading is treated as worse for this indicator.
Score composition
Stress score
Since the September 2026 release
What changed on this indicator between the two releases: new observation, source changed.
Blended from two measured components: (2/3 weight) and trend (1/3 weight); where no trend can be computed the metric is scored on its level alone. Volatility is published as a diagnostic and does not enter the blend. Full formula on the methodology page.
Level · 2/3 weight
Bounds: scored against fixed best/worst limits, used until a metric has about eight years of verified data to rank against.
Trend · 1/3 weight
Measured from the direction and size of the change between the two most recent readings.
Volatility · diagnostic, not blended
50Provisional: fewer than four historical readings are available, so this defaults to a mid range score rather than a measured figure.
Data confidence and provenance
Confidence and source
How much weight to put on the reading, the mean data-confidence across the component metrics (source quality, recency and completeness).
Source
Passenger rail performance
- Update frequency
- Quarterly
- Reliability
- high
Provenance
- Provider
- ORR
- Series ID
- trains cancelled, GB
- Frequency
- annual
- Accessed
- 1 October 2026
Known weakness
Cancellation rates are affected by industrial action and extreme weather which are temporary rather than structural; ORR methodology distinguishes planned and unplanned cancellations with different implications.
Notes
Verified official value from ORR (trains cancelled, GB), accessed 2026-10-01.
Historical values
History
| Period | Value | Relative |
|---|---|---|
| 2023 | 3.5 | |
| 2024 | 4.1 | |
| 2025 | 3.5 |