ORR Data Portal - UK Rail Statistics

The regulator's own count of its industry. The Office of Rail and Road (ORR) publishes its Accredited Official Statistics through one portal organised into nine themes - Compendia, Usage, Passenger experience, Passenger accessibility, Performance, Finance, Health and safety, Infrastructure and environment, plus the RIDDOR incident-reporting toolkit - roughly thirty named table families in all. The Usage theme alone carries the numbers the industry gets judged by: passenger journeys, kilometres and revenue on the Great Britain mainline network, broken out one column per train operating company. The headline block for January-March 2026 reads 447 million journeys, 16.3 billion passenger kilometres and GBP 2.9 billion revenue.

Two things lift this above a press-release annex. Every table ships with a signed cover sheet - source statement, revision policy, a named responsible statistician, and a legend for the [p] provisional, [b] series-break and [z] not-applicable markers - so provenance travels with the number. And the series run deep: the flagship operator table reaches back to April 2011, with older material preserved through the National Archives.

Everything below describes what Datadory delivers from that catalogue: cleaned, typed, and cut to your operators before anything ships.

Rows from the August 2026 research pass, exactly as they land before normalization:

# Table 1223 - Passenger journeys by operator, Great Britain
# annual + quarterly worksheets, April 2011 to March 2026
Time period            : 2025-26          # financial-year row label
<operator> (million)   : one column per train operating company
markers                : [p] provisional | [b] series break 2020-21 | [z] not in service

# Passenger rail usage headline block, Jan-Mar 2026
journeys               : 447 million
revenue                : GBP 2.9 billion
passenger_km           : 16.3 billion

# Cover sheet, present on every table
responsible_statistician : named per table
source_statement         : LENNON ticketing + revenue db, TOCs, TfL
notes                    : revision policy, marker legend

Three readings fall straight out of those lines. First, scale with attribution: 447 million journeys in a single quarter, each journey attributable to a named operating company rather than dissolved into a national total. Second, honesty built into the grid - the [p], [b] and [z] markers mean a parser never mistakes a provisional figure for a final one, or a structural gap for a zero. Third, the cover sheet turns citation into a one-liner: table number, statistician, source statement, done.

Six load-bearing constructs recur across virtually every table family; the theme-specific layouts - freight operations (tables numbered 1310-1365), punctuality, finance, the fares index, complaints handling - extend the same pattern and ship under additional fields on request, with column layouts confirmed with you before delivery rather than promised blind. The dictionary below is the verified core.

• Geography: the Great Britain national rail network end to end, with breakdowns at operator level, sector level and regional level - so a corridor question and a corporate question resolve against the same series.
• Temporal: long-running annual series reach back decades; the flagship operator table spans April 2011 to March 2026 across separate annual and quarterly worksheets; archived editions back to April-June 2022 remain on the portal, with older material held in the National Archives.
• Granularity: aggregated counts by operator, sector, ticket type, route or period. Never passenger-level, never service-level - this is the regulated aggregate, not a trip feed.

Set against Datadory's catalog of 1,744 datasets across 159 industries, this record scores 9/10 for quality - verified field definitions, worked markers on every column, and sample extracts confirmed on the research pass.

API, files, or your warehouse. Daily, weekly, or hourly.

You pick the channel and the cadence; the dictionary above travels unchanged through all three. Analysts building a demand panel tend toward warehouse load joined to their own geography tables; consultancies benchmarking a client operator take file extracts cut to named companies and years; product teams wiring fare-revenue features take a scoped feed of the operator set they serve. Name the operators, regions and tables when you request the sample - the sample ships first either way, and changing cadence afterward is a settings conversation, not a re-integration project.

• Operator and franchise benchmarking - place any train operating company's journeys, kilometres and revenue beside its sector and region using one consistent series instead of stitching together corporate reports that each define a passenger differently.
• Fare revenue and demand modelling - journeys, kilometres and revenue move in the same table family, so yield, average distance and volume effects separate cleanly instead of arguing through footnotes.
• Investment and infrastructure due diligence - the Finance and Infrastructure & environment themes put spending and asset context around the traffic counts, which is the combination a lender or bidder actually needs.
• Freight modal-shift tracking - the freight rail usage and performance families (tables 1310-1365) measure whether goods really move from road to rail, the claim behind most decarbonisation pitches in the sector.
• Performance and accessibility research - punctuality statistics, passenger satisfaction with complaints handling, and accessibility measures turn service-quality anecdotes into ranked, dated series.
• Journalism and academia - every figure arrives with its cover sheet: named statistician, source statement, revision policy. Citations survive scrutiny because the provenance ships with the number.

Data scientists and analysts get long, consistent panels with revision status carried as typed flags - backtesting against a series that quietly revises itself is how models learn lies. Market researchers and consultants hold the strongest claim of all: a regulated, operator-by-quarter grid is the benchmark layer every rail-side engagement ends up needing. Journalists and academics cite Accredited Official Statistics with a named responsible statistician behind each table, which is as defensible as UK transport attribution gets. Developers and builders get stable period labels and one-column-per-operator shapes that flatten into schemas without bespoke parsing per release.

It is the authoritative measurement layer for Great Britain rail - aggregates you can cite, not feeds you can query live. The Eurostat Railway Freight Transport Statistics is the natural cross-Channel counterpart for freight, and we break the trade-offs down in the head-to-head comparison. The Eurostat Passenger Transport Statistics cover rail passenger-kilometres across Europe but stop well short of operator-level GB detail. Live operational layers - vehicle locations, departures, network geometry - belong to other records such as Network Rail and TfL Open Data; teams usually pair one of those with this one, official counts for the trend, live feeds for the present.

Four honest notes from our verification pass.

Aggregates all the way down. There are no passenger-level or service-level records here. If your question needs event-level rows, this dataset anchors the trend and a different record supplies the events - we will say so up front rather than oversell.

Revisions are a feature, if you keep them. Provisional figures get revised and the pandemic era left a methodology break at 2020-21. Datadory carries [p] and [b] through as typed flags instead of silently dropping or overwriting rows, so your panel knows which numbers moved.

Theme tables ship under additional fields on request. The verified dictionary above covers the recurring core; freight, punctuality, finance, fares-index and complaints layouts get confirmed with you before delivery rather than promised blind.

Two freight tables were flagged delayed at time of research - tables 1340 and 1350, covering road haulage impact and market share - pending the April-June 2026 edition. If those two are your whole question, ask us for their status before planning around them.

Because 'the official UK rail statistics' is really thirty-odd table families wearing one badge - each with its own worksheet conventions, marker legends and cover-sheet archaeology - and relearning that layout every quarter is exactly the work nobody bills for. Datadory normalizes the catalogue to the shared dictionary above, keeps successive editions accumulating so revisions become observable history instead of overwritten files, and cuts samples to named operators, regions and tables before any recurring delivery is configured. Browse the rest of the vertical on the rail transportation data hub or the best rail transportation datasets, then request the sample - it ships first either way.