The developers of North East Thatcham say the site is “approximately 1km” from Thatcham station. A train to Reading, Newbury or London on your doorstep sounds like a good reason to build 2,500 homes here.

I’ve been collecting data on every train that calls at Thatcham since July, and digging into Network Rail’s records of why trains are late or cancelled. I use the trains myself, and most of the time they’re fine. But “most of the time” hides some real problems - and the official figures hide them even better.

Here’s what the numbers say. If you want every chart and table, they’re on the full data page.

1km? Not on foot

The “approximately 1km” is measured in a straight line from the nearest corner of the site. Nobody walks in a straight line. Routed along real roads and paths (I measured it here; also on the data page):

Most of the homes will be in the north and east. A half-hour walk each way is a long way to go to catch a train, so most new residents who use the railway will drive to it - to a station with 55 car parking spaces, most of them on the far side of the level crossing.

The level crossing

Anyone who lives in Thatcham knows the crossing. Around 9,800 vehicles use it on an average weekday, and waits of 10 to 15 minutes are common, reaching half an hour at the back of the queue (Newbury Today). The signal that controls it for trains towards Reading is over a mile away, so the barriers come down long before a train arrives.

The obvious answer is a bridge, and it’s fair to say a bridge at the station itself would be very hard to build: the railway, the River Kennet and the canal are all squeezed together there, with buildings right up against them. But a bridge further east, at Colthrop, has already been proposed. In 2019 a developer put forward 950 homes south of the railway with a road bridge from Pipers Way to Crookham Hill, so that the crossing could close - and Network Rail backed it. The council chose North East Thatcham instead, with no bridge.

The council’s masterplan FAQ says the crossing will be looked at “outside of the masterplanning process”. So: 2,500 new homes, most of them a drive from the station, and the crossing isn’t part of the plan.

So how good are the trains?

The fair answer is: better than you might think for local journeys, and noticeably worse for London.

To measure it properly I’ve looked at journeys, not just trains - Thatcham to Newbury, Reading and London, and back - and used the railway’s own standard: on time means arriving less than 5 minutes late. A journey fails if the train was cancelled at either end.

Most trains run - only about 3% of calls at Thatcham are cancelled - and journeys to Newbury and Reading mostly arrive close to time. Most delays are small: within 15 minutes, 93% to 96% of local journeys make it.

London is different. Only 65% to 70% of direct journeys to or from London arrive within 5 minutes, and 7% to 8% don’t get there at all. If you commuted to London every working day, you’d be late or stranded about one day in three.

What the official figures miss

The official performance figures make Thatcham look about as good as the rest of Great Western Railway, and better than average on cancellations. That’s because they count trains at Thatcham, not journeys. A train that leaves Thatcham on time and then terminates at Reading is a success. A train taken out of the timetable the day before was never cancelled at all.

That second one happens a lot. In September, the 09:25 from Thatcham to Paddington was planned to stop at Reading every day for a week. None of those days count as a cancellation. Across July to early November, about 1 in 6 of the regular direct trains between Thatcham and London didn’t run direct - mostly because they’d been cut back or removed in advance.

Getting home is the hard part

The journey home is the weakest of the day. From London between 16:00 and 19:00, 12% of direct trains to Thatcham failed and only 59% arrived within 5 minutes. After 21:00 there are very few direct trains home from London, so late journeys mostly mean changing at Reading.

And then there’s the last train. The last weekday train from Reading to Thatcham, around 23:00, didn’t run on 8 of 86 evenings - about 1 in 11. When that happens, it’s a taxi from Reading at around £50 to £60, or a long night.

The railway’s own compensation scheme gives a sense of scale. Under Delay Repay, which starts at 15 minutes late, a daily London commuter would be owed back about 12 single fares a year in each direction - around 5% of what they pay.

Why does it happen?

Network Rail records the cause of every delay and cancellation. For the stopping trains that serve Thatcham, from February to August 2026:

About two thirds of cancellations were down to Network Rail - mostly infrastructure faults such as points, axle counters and overhead lines, plus trespass - and about 30% to GWR, mostly train faults and crew. About 1 in 5 cancellations is recorded as “uninvestigated”. And delays on the line through Thatcham nearly doubled between February and August, mostly knock-on from problems elsewhere on the network.

In other words, much of what goes wrong at Thatcham starts somewhere else on the network.

What this means for North East Thatcham

None of this means the railway is broken. Most trains run, and for many local journeys it works well. But the case for building 2,500 homes “approximately 1km” from a station rests on the railway being a real alternative to the car, and today:

• most of the new homes will be a half-hour walk from the station, so most residents will drive to it;
• the station has 55 parking spaces, on the far side of a level crossing that already causes long queues, and which sits outside the masterplan;
• London journeys - the longest commute, and the one the railway is best placed to serve - are the least reliable, especially the journey home.

The transport work for the masterplan is still being scoped, and the draft masterplan consultation, expected late 2026 or 2027, will be the place to raise it: station parking, how residents get to the station on foot and by bike, and the level crossing all deserve proper answers before the homes are built.

Every chart, table and method note behind this article is on the Thatcham Rail Data Analysis page. Figures in the text cover July to early November 2026 unless stated; the charts update daily from Realtime Trains and Network Rail data.