Eurostat Passenger Transport Statistics (tran_hv_psmod / rail_pa)

Eurostat Passenger Transport Statistics are the Passenger Ground Transportation catalog's European benchmark record: the harmonized tables behind the question of how people actually move across the continent. The flagship table, tran_hv_psmod - Modal split of inland passenger transport - reports the share of passenger-kilometres carried by trains, by motor coaches, buses and trolley buses, and by passenger cars, for every EU member state, the EU27 aggregate and the EFTA countries, annually from 1990 through 2024.

Beside it sits the rail passenger family, led by rail_pa_total, which logs absolute magnitudes: 429,575 million passenger-kilometres across the EU27 in 2023 alone, plus passengers carried measured in thousands. Where tran_hv_psmod answers what fraction of journeys, the rail tables answer how much, in absolute passenger-kilometres. Same nomenclature, complementary jobs.

What earns the pair a place in any serious model is uniformity. Every geography reports on one methodology and one nomenclature, and rail data rest on a mandatory EU reporting framework - Regulation (EU) 2018/643, recasting Regulation EC 91/2003 - so participation is compulsory rather than volunteer-based. That is why a German share and a Greek share sit in the same chart without an asterisk farm, and why this record is the standard citation behind European modal-shift claims.

One observation per row, identifiers attached. Four rows exactly as they ship:

dataset : tran_hv_psmod
geo     : EU27_2020
unit    : PC
vehicle : CAR
time    : 2023
value   : 82.7

dataset : tran_hv_psmod
geo     : EU27_2020
unit    : PC
vehicle : TRN_BUS_TOT_AVD
time    : 2023
value   : 17.3

dataset : tran_hv_psmod
geo     : EU27_2020
unit    : PC
vehicle : BUS_TOT
time    : 2023
value   : 8.8

dataset : rail_pa_total
geo     : EU27_2020
unit    : MIO_PKM
time    : 2023
value   : 429575

Read together they sketch Europe's ground truth for 2023: passenger cars took 82.7% of EU27 inland passenger-kilometres, buses and trolley buses 8.8%, and the combined available rail-plus-bus aggregate 17.3% - leaving rail implied near 8.5 points. The fourth row converts the rail story into absolute terms: 429,575 million passenger-kilometres, roughly 430 billion, moved across the union by train in a single year. Shares for arguments, volumes for models - both in one schema.

Six fields define every observation, each definition checked against the source's own dimension structure during the August 2026 research pass. The identifying columns stack into a natural key - frequency, unit, mode, geography, year - so a question like German rail share since reunification is five filter values deep, and joins against your own country master hold up without a crosswalk table.

A handful of further columns ride along in parts of the rail family and therefore fold under additional fields on request: flag and footnote markers attached to individual observations, quarterly-frequency rows where a sibling table publishes them, spelled-out labels beside the coded values, and regional cuts carried by other tables in the rail passenger family. Name the ones your models need when you request a sample.

Geography - 37 entities in tran_hv_psmod: the EU27 member states individually, the EU27_2020 aggregate, and the EFTA countries - Norway, Switzerland, Liechtenstein and Iceland. One nomenclature codes all of them, so a 37-row panel assembles without reconciliation work.

Temporal - the modal split reaches back to 1990 and runs annually through 2024; the rail passenger tables run annually from 2004 onward. Thirty-five years of shares is enough to watch two full cycles of European transport policy land in the numbers.

Granularity - annual national observations by transport mode, a few thousand cells per table (37 geographies x 4 modes x 35 years in tran_hv_psmod), with some quarterly series appearing in related rail tables. There is no station, route or operator detail here by design; the record trades street-level resolution for the ability to compare every country on identical terms.

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

Your cadence is your call regardless of the underlying rhythm of the series - an annual-statistics corpus still lands in your warehouse on whatever beat your dashboards run, and nobody babysits a pipeline for it. Every delivery ships the complete six-field dictionary, the sample rows and the coverage profile mapped to the countries, modes and years you named.

• European market sizing - country-level rail and bus shares size the addressable base for rolling-stock, coach-building and station retail without reconciling national statistics offices.
• Modal-shift and ESG evidence - the car-share series is the number European shift claims get tested against; 82.7% for EU27 cars in 2023 settles debates a consultancy slide cannot.
• Expansion screening - rank candidate markets by public-transport share and trend back to 1990 before committing sales teams or capital.
• Infrastructure investment theses - rail passenger-kilometres from 2004 give demand-side ground truth for capacity, rolling-stock and concession models.
• Coach and long-distance bus planning - the BUS_TOT series shows where scheduled road transport already carries meaningful volume versus where it is aspirational.
• Citation-grade sourcing - stories, theses and policy briefs cite harmonized official figures rather than stitched-together national press releases.

Market researchers and consultants (relevance 3/3) get the continent's official modal-split benchmark under one nomenclature, ready to drop into any European entry or positioning study; see market researchers use cases. Investors and quant researchers (3/3) get thirty-five years of shares and two decades of rail volumes as clean demand-side series for transport and infrastructure theses; see investors quants use cases. Data scientists and ML engineers (2/3) get a compact, well-documented panel that trains and validates without cleaning marathons; see data scientists use cases. Competitive intelligence and product teams (2/3) get the official counterweight when a vendor deck claims modal-shift momentum; see competitive intel product teams use cases. Developers building data products (2/3) get a flat six-column layout that lands in a warehouse table unchanged; see developers builders use cases.

Within passenger ground transportation data, this record owns harmonized breadth: 37 countries, three modes, one nomenclature, back to 1990 - nothing else in the slice compares Europe to itself on identical terms. The neighbors own different jobs. TfL Open Data & Unified API goes metres deep inside one city network. National Transit Database (NTD) Monthly & Annual Ridership Datasets resolves US transit-agency ridership agency by agency. National Household Travel Survey (NHTS) explains why Americans travel, person-trip by person-trip. NYC TLC Trip Record Data ships trip-level microdata down to fares. Mobility Database - Global GTFS & GTFS-Realtime Feed Catalog catalogs schedules and realtime feeds worldwide. For the head-to-head between an American boarding count and this European benchmark, see CTA Ridership vs Eurostat Passenger Stats Compared.

Three things worth knowing upfront.

First, the combined aggregate is availability-weighted. TRN_BUS_TOT_AVD sums only the cells each country reports, so a missing rail year quietly shrinks the combined share. Read it as a floor and pull TRN separately wherever rail specifically matters.

Second, the design is deliberately national. No station boardings, no operator revenues, no route-level flows - city-operational work needs a different record from this catalog. The strength points the other way: 37 countries, one yardstick, thirty-five years.

Third, the field list was verified against the source's own dimension structure rather than exercised end-to-end, so the subset columns folded under additional fields on request are confirmed present in the rail family but not stress-tested during research. Say which countries, modes and years matter when you request a sample and the delivery arrives shaped to that scope.