Flight disruption data, from first-party observation
Our flight disruption data counts the delays and cancellations at a panel of airports we monitor directly, and computes what the law says each one is worth. Every number on this page comes from our own records, refreshes continuously and is free to cite with attribution.
Press enquiries: support@flightcompensations.com
€328.8M
estimated statutory compensation across observed qualifying disruptions since August 2026
1,785,091
flights tracked to date
23,652
qualifying disruptions confirmed in the last 90 days (18,268 cancellations)
Which airlines had the highest disruption rate in the last 30 days?
Carriers ranked by the share of their observed flights that met a statutory compensation threshold (3h+ arrival delay or cancellation). Minimum 20 observed flights to appear.
| Carrier | Flights observed | Qualifying disruptions | Disruption rate | Est. exposure |
|---|---|---|---|---|
| Royal Jordanian (RJ) | 22 | 2 | 9.1% | €118,500 |
| China Southern Airlines (CZ) | 24 | 2 | 8.3% | €250,000 |
| Level (LL) | 197 | 10 | 5.1% | €425,000 |
| Aeroflot (SU) | 20 | 1 | 5.0% | €76,000 |
| Thai Airways (TG) | 117 | 5 | 4.3% | €607,000 |
| Blue Bird Airways (BZ) | 25 | 1 | 4.0% | €76,000 |
| Korean Air (KE) | 50 | 2 | 4.0% | €85,000 |
| Vietnam Airlines (VN) | 53 | 2 | 3.8% | €85,000 |
| Vueling (VY) | 454 | 17 | 3.7% | €756,000 |
| Arkia (IZ) | 29 | 1 | 3.4% | €42,500 |
Which monthly releases can you cite?
Each closed month is aggregated once, after its flights have finalised, and frozen at its own URL. Quote a release when the figure has to stay true after the news cycle; the live headline numbers are for this week. Every release is also served as JSON and Markdown.
- September 2026
858,793 flights observed · 12,522 qualifying · up to €792.2M could have been owed
- August 2026
34,798 flights observed · 567 qualifying · up to €34.2M could have been owed
By airline and by airport, month by month: airline delay rankings · airport delay rankings
How do we collect this flight disruption data?
- Panel observation, not worldwide totals. We continuously monitor departures and arrivals at a panel of 160 airports across Europe, the UK, Israel and Canada, rotating coverage across the wider network. Our figures are what we directly observed — a floor on the true totals, never an extrapolation. The panel grows as our monitoring capacity does.
- A disruption counts once it qualifies. A flight enters the statistics when its verified arrival was 3+ hours late or it was cancelled — the thresholds at which EU261/UK261-family compensation can arise. Cargo and private flights are excluded.
- Exposure is a statutory ceiling, not a prediction. Estimated compensation multiplies each qualifying flight's legal band (EU261/UK261 distance bands, Canada's APPR delay tiers, Israel's Aviation Services Law) by typical seats for the aircraft or route distance. It assumes every seat is occupied and eligible — the maximum the law could put on the airline, before individual eligibility and extraordinary-circumstances defences.
- Verified, then frozen. Delays are re-verified against official flight records before a flight is treated as final; figures can therefore lag live events by a few hours. Each closed month is also frozen as a dated release that never changes after publication.
How do you cite this data?
Journalists, researchers and publishers may quote any figure on this page freely. Attribute it to FlightCompensations with a link to https://flightcompensations.com/data, and state the observation window shown beside the figure. For custom cuts — a specific airline, airport, route or period — write to support@flightcompensations.com and we’ll run the numbers from the same records, with the same methodology.
What does this flight disruption data measure?
We watch a fixed panel of airports and record what happened to the flights through them: the schedule, the actual arrival, and whether the airline cancelled. From that we compute the statutory ceiling each disruption carries, so the totals read as money rather than as counts. The panel is a sample and we say so on every figure.
Nothing here is modelled or projected. If a number is not in our records, it is not on this page.
How do these flight delay statistics put a price on a delay?
Most published delay figures count events. Ours count entitlement. A three-hour arrival delay on a short hop and the same delay on a long-haul route are one statistic apiece and €250 against €600 in law, and that gap is the thing worth reporting.
| Flight distance | EU261 amount per passenger for a 3+ hour arrival delay |
|---|---|
| Up to 1,500 km | €250 |
| 1,500 to 3,500 km, and flights inside the EU over 1,500 km | €400 |
| Over 3,500 km | €600 |
The live radar shows the same observations on a map, and the EU261 guide explains the bands the money comes from.
What does the cancellation data leave out?
A cancellation appears here as soon as the airline files it. What never appears is the reason, because no data feed carries one. That matters for any story built on these numbers: the totals show what was at stake, not what was owed. The airline can still show a storm closed the airport.
Readers who want their own flight rather than the aggregate should use the flight check, which reads that one flight’s record instead of the panel.
What do journalists ask about this flight disruption data?
Where does this flight disruption data come from?
From our own monitoring, not from a regulator's release. We watch a defined panel of airports, record scheduled and actual times for the flights that run through them, and compute what each qualifying disruption is worth. Every figure on the page is generated from those records when the page loads.
Can I use these figures in an article?
Yes, free of charge. Name us as the source, link to this page, and state the observation window shown beside the number you quote. If you need a cut we do not publish, such as one airline or one airport over a set period, write to the press address above.
Are these numbers a worldwide total?
No, and we never present them that way. They describe a panel of monitored airports, which is a sample rather than the whole sky. A figure here is a floor for what happened at those airports, not an estimate of European or global traffic.
How often does the cancellation data update?
Continuously, with the page itself rebuilt hourly. The underlying sweeps run every fifteen to thirty minutes on the busiest airports, so a disruption reaches the totals within an hour of the aircraft landing or the airline filing the cancellation.
Sources
- Regulation (EC) No 261/2004 — the regulation itself, in every EU language
- UK Civil Aviation Authority — passenger rights — the UK regulator's own guidance and complaint route