Which UK bus operators run on time?

We continuously compare live bus GPS positions against the published timetable and record whether each bus was on time (within 5 minutes). These league tables cover local bus services over the last 30 days — 201,006 observations. No other UK site publishes this.

These are our own measurements, not official statistics. They are not comparable with the Department for Transport's figures further down this page and read roughly 25–30 points lower by construction. Why, and what else they don't tell you →

Operators ranked by punctuality

Colour shows where an operator sits against the others measured here — top third green, middle amber, bottom third red — not a pass mark. Every figure in this table is low against the official ones because it is measured differently (why), so compare operators with each other and never with a Department for Transport percentage. An operator appears once, under the name it trades as, even where it holds several licences.

Official punctuality by local authority (2025)

The Department for Transport publishes its own punctuality figures each year. They cover a whole council area rather than individual routes, but they go back to 2005 — far further than any live tracking. "On time" here means between 1 minute early and 5 minutes 59 seconds late.

Excess waiting time applies to frequent services only, so it is blank for many areas. Source: DfT bus reliability and punctuality (BUS09), Open Government Licence v3.0.

By English region (2025)

Each figure covers every council in the region combined, so it sets the context the table above sits in rather than competing with it.

How this is measured

Operators publish their buses' live GPS positions through the DfT Bus Open Data Service. We match each tracked bus to its scheduled journey, estimate its delay from its position along the route, and store an anonymous observation. "On time" follows the industry convention: departing between 1 minute early and 5 minutes late. Services need 50+ observations (operators 200+) in the last 30 days to be ranked, so a single bad day doesn't condemn a service. See a route's own page for its individual score.

Local bus services only. That 1-early/5-late window is the Department for Transport's standard for a registered local bus service, and it is the only thing it measures fairly. A long-distance coach has timing points hours apart and runs on motorways; scoring it against a bus yardstick says nothing about the coach and a great deal about the yardstick. Coaches, trams, ferries and rail replacement are still tracked and still appear on their own pages — showing average delay, which needs no window — but they are not ranked here.

The unit is a service, not a timetable row. One service often exists as several routes in the raw data — a school working beside the all-day one, or the same line split by direction — and those ids are reissued when an operator republishes its dataset. Observations are gathered onto the operator's licence code, line and region instead, which is the same identity the weekly service history uses and survives a rebuild.

What these figures don't tell you

This is not a complete record of every bus that ran. We see a journey only when the bus is broadcasting its position and we can match that position to a scheduled journey, and buses don't always track — equipment fails, drivers sign on late or to the wrong journey, and some operators publish no live positions at all. If a route or an operator isn't listed here, it does not mean their buses run badly. It means we haven't seen enough of them to say anything.

Seven things worth knowing before quoting one of these numbers:

• A bus that never ran is invisible, not late. A cancelled journey, or one that ran with its tracker off, produces no observation at all — so it never counts against a score here. Punctuality and reliability are different questions, and this page answers the first.
• The worst delays drop out. When a vehicle's position implies it is more than 45 minutes behind schedule (or 8 minutes ahead), we discard the match rather than record a number we don't trust — at that distance it is as likely to be the wrong journey. The honest consequence is that severe delays are under-counted and these scores lean optimistic.
• Delay is estimated from position, not measured at a stop. We work out where along its route the bus has reached, interpolate the time its timetable gives for that point, and compare the two. A bus more than 3 kilometres off the route is treated as a bus we have mis-identified rather than a bus running late, and is discarded. That is a fair estimate of how late a journey is running; it is not a record of when it actually left any particular stop.
• Coverage follows attention, not need. Observations are recorded when someone opens a stop page, plus an automated daytime sweep of one busy stop in each of 25 towns. Busy urban routes are sampled far more heavily than rural ones, and night buses barely at all.
• Small samples move a long way. A service qualifies at 50 observations across 30 days, which can be a few journeys on a few days. One afternoon of roadworks can move a score by ten points. The observation count is printed beside every figure for exactly this reason — check it before believing a surprising one.
• An operator table is not a like-for-like comparison. Operators are ranked on the routes we happened to observe, which is not the same as the routes they run. A company working city-centre services in congestion will place below one running interurban routes on open roads, whatever either of them does well. It measures the conditions the buses run in at least as much as the company running them.
• These are not the DfT's numbers and are far lower. The official figures further down this page put many councils in the eighties and nineties; nothing here reaches 70%. That gap is the method, not a disagreement about which buses were late. The DfT measures a departure, at a timing point, against the time printed for that stop. We measure a bus's position, at an arbitrary moment, against where its schedule implies it should have got to — and a bus that leaves every timing point punctually still spends much of its run a few minutes off that line. Ours will always read lower. Compare our operators with each other; compare the DfT's councils with each other; never compare across the two.

The Department for Transport figures further up are collected on a different basis, at stops rather than along the route, over a full year rather than 30 days. They are here for the long trend and the decade of history we can never collect ourselves — not as a check on our numbers, which measure a different thing and sit roughly 25–30 points below them by construction.

Live position data: DfT Bus Open Data Service (Open Government Licence v3.0). Observations are our own derived measurements and carry no personal data.