The UK's most delayed stations — where late trains come from

In short: London King's Cross emits the most delay (5.8 min average for late services) and absorbs the second-most (6.8 min). Edinburgh Waverley emits 5.1 min but absorbs 7.2 min — the long-distance routes accumulate delay en route. The stations that emit and the stations that absorb are mostly the same big terminals, but the amount differs depending on whether they're the start or end of the journey.

The UK's most delayed stations

Every late train has an origin and a destination. Some stations are net delay emitters — they start more late trains than they absorb. Others are net delay absorbers — they receive late trains and pass them on to the next leg. The two lists look similar at first glance (the same big terminals appear in both), but the amounts differ.

---

Two kinds of station delay

Emit delay is delay that starts at a station. If the 08:00 from King's Cross to Edinburgh leaves 6 minutes late, every following stop on that route is at least 6 minutes late, and the delay cascades for the rest of the journey. King's Cross in this case is the source — King's Cross "emits" the delay.

Absorb delay is delay that ends at a station. If the 14:00 from Edinburgh to King's Cross arrives 12 minutes late, King's Cross is where the delay lands — King's Cross "absorbs" the delay (and then re-emits it on the next departure if there's a tight turnaround).

The same station can both emit and absorb — in fact, every major terminal does both, because every departure is somebody's origin and every arrival is somebody's destination. The interesting question is the net flow: is a station a bigger source than a sink, or vice versa?

For a station like King's Cross:

• It emits delay because it sits at the edge of the network's physical capacity (6 platforms, lots of long-distance departures per hour, complex throat).
• It absorbs delay because long-distance services from Edinburgh, Leeds, and Newcastle terminate there, carrying accumulated delay from the East Coast Main Line.

The two effects compound: a late arrival at King's Cross makes the next departure late, which makes the next arrival at Edinburgh late, which makes the next departure from Edinburgh late, and so on.

---

The stations that emit the most delay

The chart below ranks the top 20 UK stations by average delay for services that departed late, over the last 30 days. Sample threshold: 200 services minimum.

London Euston 5.4 min, Edinburgh Waverley 5.1 min, Manchester Piccadilly 4.9 min, Glasgow Central 4.7 min, Birmingham New Street 4.5 min lead the chart.">

Rank	Station	Emit delay
1	London King's Cross	5.8 min
2	London Euston	5.4 min
3	Edinburgh Waverley	5.1 min
4	Manchester Piccadilly	4.9 min
5	Glasgow Central	4.7 min
6	Birmingham New Street	4.5 min
7	Leeds	4.4 min
8	Bristol Temple Meads	4.2 min
9	London Paddington	4.0 min
10	Reading	3.9 min
11	Sheffield	3.8 min
12	Doncaster	3.7 min
13	York	3.6 min
14	Cardiff Central	3.4 min
15	Newcastle	3.3 min
16	London Liverpool Street	3.2 min
17	Brighton	3.1 min
18	London Victoria	3.0 min
19	London Waterloo	2.9 min
20	London Bridge	2.8 min

Three clusters are visible:

• The "edge of network" terminals. King's Cross, Euston, Paddington, Liverpool Street, Victoria, Waterloo, London Bridge — the seven big London terminals. Each sits at the end of a main line, with constrained throat layouts and high platform density.
• The long-distance origin cities. Edinburgh, Manchester, Glasgow, Birmingham, Leeds, Bristol, Cardiff — these are the start points for intercity services. They're net emitters because every long-distance departure begins there.
• The mid-route junctions. Doncaster, York, Newcastle, Reading, Sheffield — these are the through-stations where one network meets another. They're emitters because every departure through them depends on the inbound service being on time.

London terminals feature heavily because every long-distance service in the UK either starts or ends in London (with a few open-access exceptions). When you compress all the long-distance origin city delay into the London terminals, they dominate the ranking.

---

The stations that absorb the most delay

The same 20 stations appear when you flip the analysis to look at arrival delay, but with different amounts.

Edinburgh Waverley 7.2 min, London King's Cross 6.8 min, Manchester Piccadilly 6.4 min, Glasgow Central 6.2 min, London Euston 5.9 min lead the chart.">

Rank	Station	Absorb delay
1	Edinburgh Waverley	7.2 min
2	London King's Cross	6.8 min
3	Manchester Piccadilly	6.4 min
4	Glasgow Central	6.2 min
5	London Euston	5.9 min
6	Leeds	5.6 min
7	Birmingham New Street	5.4 min
8	Stansted Airport	5.3 min
9	Liverpool Lime Street	5.0 min
10	Bristol Temple Meads	4.9 min
11	Sheffield	4.7 min
12	Newcastle	4.6 min
13	London Paddington	4.5 min
14	York	4.4 min
15	Reading	4.3 min
16	Doncaster	4.1 min
17	London Liverpool Street	4.0 min
18	Brighton	3.9 min
19	London Bridge	3.8 min
20	London Waterloo	3.7 min

The arrival delays are systematically higher than the departure delays — by an average of 1.4 minutes per station. That's the accumulation effect: a 5-minute slip at origin is a 7-minute slip at destination on a 4-hour route. The terminal stations (Edinburgh, King's Cross, Manchester, Glasgow, Euston) absorb the most because long-distance routes terminate there.

A few stations move up or down between the two rankings:

• Edinburgh Waverley is the biggest mover — from #3 emitter (5.1 min) to #1 absorber (7.2 min). Edinburgh is the northern end of the East Coast Main Line; long-distance services from King's Cross accumulate 4 hours of delay before arriving at Edinburgh, but the departure side has stock that's been on the network since 06:00 with fewer compounding issues.
• Stansted Airport appears in the absorber list (5.3 min) but not in the emitter list. Stansted is a destination for many long-distance services that have crossed half the country to get there — Greater Anglia's airport service, CrossCountry from the south and west. Few services depart Stansted that accumulate much delay before arrival at the next stop.
• London terminals appear high in both lists because every long-distance service either starts or ends at a London terminal.

---

Why these stations are bottlenecks

Three structural reasons these specific stations sit at the top of both rankings:

Physical constraints. King's Cross has 6 platforms for long-distance services. Euston has 18 platforms but a constrained throat — the approach tracks merge before reaching the platforms, so a late arrival at the throat delays every following service. Paddington has 12 platforms but only 6 are usable for long-distance at peak. These terminals are at the edge of physical capacity; adding trains is impossible without rebuilding.

Network convergence. Every UK long-distance route starts or ends in London (with the small exceptions of CrossCountry's regional routes). When you concentrate all those services through a handful of terminals, the terminals become system bottlenecks — every upstream delay converges there.

Junction topology. Reading, Doncaster, York, and Leeds are all major junctions where multiple main lines meet. A delay entering the junction cascades through every service that uses the junction in the next 30 minutes. These mid-route junctions don't have the same volume as London terminals, but they're structurally fragile.

The combination is what makes them bottlenecks. A station can have low volume but high fragility (a junction with three converging routes and one platform conflict can stop a main line for an hour). A station can have high volume but low fragility (Waterloo has 24 platforms; a single platform conflict absorbs quickly). The stations at the top of these rankings have both — high volume and high fragility.

---

What this means for your commute

If your commute passes through one of these stations, plan around the structural bottleneck:

• Through King's Cross or Euston: allow buffer time for the first leg of any onward connection. A late arrival at King's Cross is very likely; a connection at 10 minutes is risky. Plan for 15–20 minutes between trains.
• Through Edinburgh, Manchester, Glasgow, or Birmingham: these are the big mid-route destinations. If your journey terminates here, expect the arrival to be later than the in-seat announcement suggested (compounding delay from 3+ hours of running). If your journey continues, allow buffer for the cross-platform interchange.
• Through Reading, Doncaster, York, or Leeds: these junctions are where one main line meets another. A delay on one route propagates to all routes using the junction. Check the live board (or wherever you're starting) before you leave — if Reading is delayed, your connection through Reading is delayed.

For long-distance services, the reliability ranking by route is a better guide than the operator ranking — the route-level number tells you what to expect on your specific service.

For Delay Repay purposes, the absorb-delay chart tells you how much delay to expect on arrival at your destination. If you're travelling into Edinburgh or King's Cross, allow extra margin and expect to claim more often. The how-to-claim guide covers the process.

---

Methodology

The figures above are 30-day rolling averages of per-day station-level delay, joined from the same data sources as our mainline ranking. The emit-delay figure is the average departure delay across all late services at that station; the absorb-delay figure is the average arrival delay across all late services terminating at that station.

• Window: rolling 30 days, ending yesterday.
• Late services only: on-time and cancelled services don't contribute to the average.
• Station definition: origin/destination CRS code, joined at the calling-point level. Each station appears once as origin and once as destination in the data.
• Sample size threshold: at least 200 services in the window. Quiet rural stations are excluded from the top 20.

The colour bands in both charts are a five-step gradient (green → red) by delay minutes. Stations near 7 minutes are red; stations near 3 minutes are green.

---

Last verified: September 2026. Numbers come from the last 30 days of train running data on the National Rail network and refresh nightly. If you spot something that looks wrong, drop us a note.