Slow Travel
Night trains are back, and the economics are still difficult
Sleeper services are expanding across Europe for the first time in decades. The operating maths that killed them has not gone away.
The theory of night trains is well covered elsewhere. This is about the version you meet in practice.
What holds up in practice
- A sleeper carriage carries a fraction of the passengers of a day coach for the same path.
- Rolling stock is the binding constraint — almost none has been built for decades.
- Cross-border operation multiplies the regulatory cost of every route.
The revenue per carriage is the core problem
A day coach seats around eighty people; a couchette carriage sleeps perhaps forty and a sleeper considerably fewer. The train occupies the same path, needs the same crew and burns similar energy, so revenue per kilometre is structurally lower.
Ticket prices therefore have to sit well above a seat fare to break even, which puts sleepers into competition with budget airlines rather than with day trains. Track access is charged per train rather than per passenger, and an overnight path competes with freight and with the engineering possessions most networks schedule at night, so the slot itself is both expensive and awkward to obtain.
Rolling stock barely exists
Most European sleeper carriages in service were built in the 1970s and 1980s, and refurbishment can only extend that so far. New builds have restarted but in small numbers, with long lead times and high unit costs because the production runs are tiny. This is why expansion announcements are usually about routes rather than frequency: there are not enough carriages to run more trains.
New builds also have to satisfy accessibility and crashworthiness standards the 1970s stock never faced, which constrains compartment layout and pushes up the cost per berth in a way refurbishing an old carriage does not.
Crossing borders costs more than distance does
Every additional country adds signalling systems, safety authorisations, driver certification and often a locomotive change. A route through four countries can require approval processes in all four, each on its own timetable.
This is the single largest reason promising routes stay unlaunched long after somebody has announced them. Crews are part of the same problem, since drivers have to be certified for the specific route and frequently for the language used on it, so every border adds a crew change, a depot and a rostering puzzle.
The demand is real and specific
Sleeper demand is strongest where the alternative is an unpleasant early flight and the journey is between six and twelve hours. Below that, a day train wins; above it, most travellers revert to flying regardless of preference. Business travel has largely not returned to sleepers, which removes the highest-yield segment operators used to rely on.
In practice, what remains is strongly seasonal, peaking in summer and around ski weeks, and a fleet sized for the peak sits idle for months, which is precisely the wrong shape for an asset this expensive to buy.
What actually makes a journey good
Private compartments and reliable air conditioning matter more to repeat custom than anything else on board. Arrival time is the hidden variable: a service that arrives at 05:40 is much less useful than one arriving at 08:30, regardless of comfort.
The detail that matters: the best-run services build in a slow final section precisely to land at a civilised hour. Many overnight services are also split and joined en route so that portions can serve different cities, and being shunted at two in the morning is the most reliable way to be woken, which is worth checking before choosing which berth you book.
That varies by region, and it is worth checking locally.
How to book them without frustration
Sleeper inventory opens later than day trains on many networks and sells out unevenly, with the cheapest berths going months ahead. Booking directly with the operating railway is usually cheaper and always easier to change than going through an aggregator.
Where a route is operated jointly, the same berth can appear at different prices on different national sites. Rail passes are the other trap, because a pass covers the journey and not the bed: the berth reservation is a separate and strictly limited item, and on popular sleepers those go long before anybody planning around a pass thinks to look.
The takeaway
Book early, choose the arrival time before the price, and expect the route map to keep changing.
What holds up over a week outdoors is rarely what looks best at the trailhead.
Questions readers ask
Are night trains actually cheaper than flying?
Rarely on the headline fare. They can be competitive once you count airport transfers, baggage, and a night of accommodation you no longer need.
Is it worth paying for a private compartment?
If you need to sleep properly, yes. The difference between a shared couchette and a private berth is the difference between arriving rested and arriving having lain awake.
