Live rail map & route planner

Include junctions & non-station locations (to force a particular route)

Waypoints (0)

Add two or more stops. If no continuous rail path exists between consecutive stops, the result is split into separate routes.

Plan a route over the real railway

This traces the track a train would actually run over between the points you give it — not a straight line, and not a road route. The result is real geometry: the paths through junctions, the side of the station it arrives on, and the line it takes where two run parallel.

Three ways to ask

Give it stations by name, CRS code or TIPLOC and it will route between them in order. Give it coordinates and it snaps each one to the nearest reachable track, which is how you route to a depot, a junction or a spot with no station at all. Or give it a planning ID — a train's own schedule identifier — and it follows that service's booked calling pattern.

You can also click points straight onto the map. Waypoints are visited in the order you set them, so the order changes the answer wherever the network offers more than one way round.

What comes back on each leg

Every leg reports the Engineer's Line References it runs over with the distance on each, the gradient in the direction of travel, the line speed and electrification where the survey records them, and any temporary speed restriction whose extent the leg crosses. Distances are measured from the geometry itself rather than derived from the routing cost, so a diverted or roundabout route reports the length it really is.

Guide: what an ELR actually is — the codes on every leg, why no official source names them, and why the letters cannot be decoded (MLN1 is the Great Western, not the Midland).

Why one request can return several routes

Not every list of places can be joined by one train. Where a continuous path does not exist — or would mean running the wrong way down a line a train cannot use — the answer comes back as more than one route, which is the honest representation of a journey that needs a change. Routes built from a planning ID prefer a single continuous run visiting every calling point; routes you build by hand are allowed to split.

Sidings, yards and crossovers are all in the network, but a quarter of it is not running line. Service track is costed several times its length so a route is not planned through a depot to save a few hundred metres, while crossovers stay cheap because they are the legitimate way between adjacent running lines.

Finding the train first

The plan tab searches the working timetable rather than this map: an operator and a headcode, a time and a station, or a unit number all narrow it down, and the result can be routed with one click.

Guide: how to find one train among half a million — what a headcode, a UID and a planning ID each are, and which one you actually have.

The dots, and what they are not

Live positions come from the signalling system rather than from GPS. A train's headcode sits in a berth — roughly the section of line between one signal and the next — and steps into the following berth as the train passes. That step is the entire basis of the map: no coordinates, no speed, no heading.

Two consequences follow, and both look like faults until you know. A berth is an area rather than a point, so the dot is accurate to a signal section and no better. And because many neighbouring berths share one location reference, roughly half of all reported movements do not move the dot at all — a train can report four times in a row and appear frozen.

A dot's fill is its operator and its ring is how late it is. The ring has a third colour for unknown, deliberately not green: around 13% of what moves here is freight, which the passenger forecast feed does not cover, so we know where those trains are and nothing about their punctuality. Calling that “on time” would be inventing the most interesting fact about them.

Fainter dots are inferences. Where the describer cannot see a train — some lines have no mapped berths whatsoever — it is placed on its booked path shifted by how late it is running, and drawn differently because an inference should not carry a report's confidence.

Guide: where the dots come from, and when the dots stop, and what that means — including what happens when an entire describer area fails.

Why the drawn route is not the timetable

A route here is computed over a model of the physical network, not looked up. The timetable knows a train calls at Sheffield; it does not know which line it takes to get there. So the track is joined on afterwards from mapped geometry, and the answer is an inference — a good one, but one that can differ from what the train actually did on the day.

The model knows more than track shape. About a quarter of Britain's railway ways are sidings, yards, crossovers and spurs rather than running lines, and a planner that ignores that will happily route a train through a depot because depots are well connected. Service track is charged six times its real length here — a last resort rather than an impossibility, because a station's own platform road is sometimes tagged that way.

Distances are measured from the drawn geometry rather than derived from the routing cost. Once a number has been weighted for any purpose it stops being a measurement, and a siding route would otherwise report six times its true length.

🚉 Live departures & arrivals →

TransitRadar is an independent site, not affiliated with, endorsed by or operated by National Rail, Network Rail, or any train or bus operator.

Live trains Live buses Disruptions

Live UK rail map and route planner