For anyone travelling: check National Rail or your operator for the current position. Nothing on this page is travel advice, and the closures around Lewes will change as the line is recovered.
Where it happened
ITV Meridian reported the site as between Landport Road and the River Ouse, roughly a mile north of Lewes railway station. BBC Verify placed it there independently, by matching features in a photograph of people standing on an overturned carriage, including the treeline on the left of the shot. The BBC separately understands the train left the tracks in the Landport area of Lewes.
A mile north of the station puts it on the approach from Cooksbridge and Haywards Heath, not in the platforms. That matters for reading the running data below: the last position report is recorded at Lewes, which in signalling terms means the train had reached the Lewes area, not that it had come to a stand in the station.
What the running data recorded
The service involved was the 14:24 London Victoria to Eastbourne, headcode 1F34. Its run is public: it is in National Rail's Darwin feed, the same source that drives station departure boards and this map. Reported times, all from that feed:
| 14:24 | departs London Victoria |
|---|---|
| 14:29 | Clapham Junction |
| 14:41 | East Croydon |
| 15:12 | Gatwick Airport |
| 15:33 | Haywards Heath |
| 15:45 | Cooksbridge |
| 15:51 | Lewes. The last confirmed report of any kind |
Nothing after 15:51 is a real observation. The feed carried one further entry, a departure from Lewes timed 15:53 that was published half an hour later, and no signalling report corroborates it. Estimates for the stations ahead kept being reissued until 17:35, when the remaining calls were cancelled and the service stopped being updated.
What this map got wrong
Between stations, this map does not know where a train is. It works the position out from the last place the train reported and the time it is due at the next one, and draws it along the real track in between. That is fine for a train that is running. It is badly wrong for one that has stopped, because the railway's estimates for the stations ahead keep being pushed back for as long as the train stays put, and the calculation follows them.
So for about an hour and three quarters after 15:51, this map showed the train continuing down the line towards Eastbourne, several miles past where it actually was, carrying a speed of around 68 mph. That is our defect, not the data's.
It is now fixed. When a train's last real report goes stale, the map stops calculating and leaves it where it was last actually seen, and the train's card says Not reporting with the place and time of that last report. Measured across a full day, about one train in fifty is in that state at any moment, mostly during disruption. There is a note in the changelog, and the build log covers how the positions are worked out.
Sources
Location and confirmation: ITV Meridian and BBC News, including BBC Verify's photograph analysis. Running times: National Rail's Darwin real-time feed, which this site consumes under licence and which is the source of every train shown on the map.
Last updated 13 August 2026. This is a record of what the running data showed on the day and what we changed as a result, not a running account of the incident.