Transport for Wales Performance

Independent performance data for Transport for Wales from Network Rail TRUST train movement records.

Based on 84 days of data (2026-02-27 to 2026-05-22) | Last updated: 22 May 2026
Reliability
80.8%
ran and arrived on time
On-time (if ran)
92.5%
within 5 minutes
Cancellation Rate
12.6%
of services
Avg Delay
1.6 min
minutes late
Compared to the previous 60 days: On-time performance has declined by 0.3% compared to the preceding period.

Performance by time of day

Period On-time % Avg Delay Services
Morning peak96.2%1 min144,445
Off-peak93.4%1.4 min377,854
Evening peak91.3%1.7 min167,462
Late/early89.5%2.1 min233,374

Performance by day of week

Day On-time % Avg Delay Services
Monday93.1%1.5 min140,609
Tuesday92.4%1.6 min143,218
Wednesday92.1%1.6 min142,186
Thursday93%1.5 min143,516
Friday92.1%1.6 min139,328
Saturday91.3%1.8 min145,323
Sunday94.4%1.3 min68,955

Monthly history

Month On-time % Avg Delay Cancellation Rate Coverage
April 202691.7%1.7 min11.1%100%
March 202692.7%1.5 min16.8%87.1%

Best performing routes

Route On-time % Avg Delay Services
Machynlleth to Aberystwyth100%0.5 min285
XRS to Rhymney100%0 min270
XCC to Aberdare100%0 min128
XTY to Pontypridd100%0.1 min69
XCC to Merthyr Tydfil100%0 min62

Worst performing routes

Route On-time % Avg Delay Services
Penarth to Heath Low Level0%12.5 min10
Ebbw Vale Town to Tondu5.6%18.4 min18
Merthyr Tydfil to Fernhill8.3%14.6 min12
Llandudno to Manchester Piccadilly19%16.6 min21
Pontypridd to Cardiff Queen Street26.7%7.3 min30
View Transport for Wales Delay Repay guide →

How we calculate these figures

These statistics come from Network Rail's TRUST system, which records the actual time every train passes through each point on the network. We receive this data in real time and calculate performance independently.

Reliability is our headline metric. It answers the question most passengers care about: "If I turn up for a scheduled train, what are the chances it runs and gets me there on time?" It's calculated as: (trains that ran and arrived within 5 minutes) divided by (total scheduled trains). A cancelled train counts against reliability, because you can't ride a train that doesn't exist.

On-time (if ran) shows how punctual trains were, but only counting trains that actually ran. This is useful for understanding whether delays are the main problem, or cancellations. If reliability is low but on-time is high, the operator's main issue is cancellations rather than lateness.

Cancellation rate is the percentage of scheduled services that were cancelled, including both full and part-cancellations.

Average delay is the mean delay across all arrivals. Trains that arrived on time count as 0 minutes delay. This tells you how late trains typically are when they don't run to time.

What we exclude: We don't count "off-route" movements — these are signals recorded when a train passes through a station that isn't on its scheduled route (for example, a Heathrow Express triggering a sensor at a Great Western station it passes through). Including these would unfairly lower an operator's on-time score.

How this differs from official figures: The Office of Rail and Road (ORR) publishes official statistics using different thresholds and reporting periods, with post-publication corrections. Our figures are independently calculated and not revised after publication. For official statistics, see the ORR Data Portal.

Full methodology details →

Data source: Network Rail TRUST train movement data. Contains public sector information licensed under the Open Government Licence v3.0. These are independent calculations and are not official statistics published by the Office of Rail and Road. On-time is defined as arriving within 5 minutes of the planned time. For official statistics, visit the ORR Data Portal.