How Much Time to Leave for a Train Connection (And When the Minimum Isn't Enough)

The official minimum connection time printed on your itinerary is the shortest a connection can legally be sold at — not a recommendation for how much time you should actually leave.

Same-station connections are typical across most of Europe and Japan's shinkansen network. Cross-station connections requiring a taxi, metro or long walk between two separate terminals are more common in parts of the USA and in some Southeast Asian cities where rail and long-distance bus terminals sit apart from each other — factor real transfer distance into your buffer, not just the timetable gap.

Every rail itinerary with a connection carries a printed "minimum connection time" — the shortest gap the booking system will allow between an arriving and departing train at that station. It is tempting to treat that number as safe, since the system itself won't sell you anything shorter. It is not a safety margin; it is a floor, calculated under the assumption that your first train arrives exactly on time, on the expected platform, and that you know precisely where you're going next. Real journeys deviate from all three assumptions often enough that building in your own buffer is worth the planning effort.

What the minimum connection time actually accounts for

Operators calculate this figure from walking distance between the likeliest platforms at that station, average passenger walking speed, and sometimes a small contingency. It generally does not account for: the first train running late, an unexpected platform change announced at the last minute, needing to find a ticket machine or information desk, navigating an unfamiliar station with luggage, or a connecting train that itself departs a few minutes early from a different platform than expected. Any one of these is common; several stacking together on a single journey is not rare.

A practical buffer, by connection type

• Same-station, same-platform-area connection, no border, familiar transit hub: minimum connection time plus 15-20 minutes is usually comfortable.
• Same station but different platform zones, or a large multi-level terminal: add 25-40 minutes over the stated minimum, more with heavy luggage or reduced mobility.
• Cross-station connection (a different train station entirely, requiring a taxi, metro or a walk of more than a few minutes): treat the minimum connection time as close to meaningless and budget the actual transfer time between the two stations, plus 30-45 minutes of contingency, especially in an unfamiliar city.
• International or border connection with a passport check, customs formality, or a change of rail gauge/system: add at least 45-60 minutes beyond the stated minimum, more during peak holiday travel periods when border queues lengthen.
• Last connection of the day on any route: always add extra buffer regardless of type, since missing it can mean an overnight stay rather than a later same-day departure.

What actually happens if you miss a connection

The outcome depends heavily on how the tickets were booked, and this is worth understanding before it happens to you, not during:

One combined ticket, sold as a single through-journey

If both legs were booked together as one itinerary by the same operator or through a booking system that explicitly protects the connection, a missed connection caused by the first train's delay is usually the operator's problem to fix — typically rebooking you onto the next available departure at no extra charge, sometimes with a delay-compensation claim available afterward.

Two separately-booked tickets treated as unrelated

If you booked two independent tickets — even for the same journey, even from the same operator's website but as separate transactions — a missed connection is generally treated as your own missed departure. The first ticket being late is not the second ticket's problem in the eyes of the system, and you may need to buy a new onward ticket at whatever fare is available, which can be considerably more expensive than what you originally paid.

What to do in the moment

Go to a staffed ticket window or information desk at the station immediately rather than trying to resolve it from an app — staff can see real-time availability and rebooking options that consumer apps often cannot, and can advise whether your specific ticket type entitles you to a free rebook.

Reducing the risk before you even book

• Book connecting legs as one itinerary wherever the option exists, even if it costs marginally more than booking separately — the protection is often worth it.
• Check whether the connection is same-station or cross-station before assuming; station names on a booking screen do not always make this obvious, and a quick map check resolves it in seconds.
• Avoid scheduling the last possible connection of the day if there is any way to build in a slightly earlier departure instead — the cost of an earlier train is usually far less than the cost of an unplanned overnight stay.
• Travel light through connection-heavy itineraries. A rolling suitcase and one bag moves through an unfamiliar station meaningfully faster than three separate items and an awkward hand-carry.

The honest bottom line

No buffer eliminates risk entirely — delays and disruptions happen on every rail network in the world. What a sensible buffer does is convert "missed connection" from a likely disaster (an unplanned night in a station city, an expensive walk-up fare) into a minor inconvenience (a later train, a longer wait) far more often than the bare system minimum ever will.

Reading a station's layout before you rely on a tight connection

Large interchange stations are frequently built across multiple levels, with platforms for different services grouped in physically distant zones connected by long corridors, escalators or even a short walk outside and back in. A "12-minute" connection at a station with platforms 1-6 clustered together is a very different proposition from a "12-minute" connection where your arriving platform is at one end of a half-kilometre concourse and your departing platform is at the other. Where possible, check a station map in advance for any connection under 20 minutes, rather than assuming all platforms at a named station are close together — this single check catches a meaningful share of connections that look fine on paper and aren't.

What conductors and staff can and can't do for you

If you're clearly going to miss a connection because the first train is running late, on-board staff on many networks can radio ahead in some circumstances, but this is far from universal and should never be assumed. It is generally more productive to ask on-board staff, as soon as a delay becomes apparent, what your options are at the connecting station specifically, rather than waiting until you've physically missed the connection to start asking. Staff who know a delay is developing can sometimes advise an alternative connection before you even arrive, which is far more useful than troubleshooting after the fact.

Self-connections vs protected connections, revisited

It's worth restating plainly: a "self-connection" — two tickets you assembled yourself because no single booking covered the whole route — carries no institutional obligation for either operator to look after you if the first train runs late. This is different from a through-itinerary sold as one product, where a delay on the first leg is explicitly the selling operator's problem to solve. When comparing a slightly cheaper self-connected itinerary against a slightly pricier protected one, factor in not just the ticket-price difference but the realistic cost of the self-connection failing — a same-day walk-up fare on the second leg, or worse, an unplanned overnight stay.

General information for planning purposes, not travel advice. Fares, schedules, reservation rules and passenger rights vary by country and operator and change over time — confirm current details directly with the operator or booking platform before you travel.