Edition 1 · June 2025 to May 2026 · US federal data

The JFK Delay Index

We analysed 102,858 US domestic departures from JFK across twelve months of federal on-time data. Three things came out of it, and two of them contradict what people believe about this airport.

Departures analysed

102,858

Delayed 15+ min

21.4%

Cancelled

3.1%

Mean delay

17.9 min

What time of day is JFK most delayed?

JFK is most delayed late in the evening, and the effect is enormous. A departure scheduled in the 22:00 hour is delayed 15 minutes or more 37.7% of the time. One scheduled at 7:00 is delayed 9.5% of the time. That is a 4.0-fold difference in your odds, decided entirely by which flight you book.

The curve climbs almost without interruption from first light to last departure, which is what a day of accumulating knock-on delay looks like. It also explains the next finding, because the thing that accumulates is not weather.

What actually causes JFK delays?

JFK delays are caused overwhelmingly by the airlines, not by the sky. Weather accounts for 3.5% of delay minutes. Late-arriving aircraft and the carriers' own operations together account for 68%.

One caveat we will not bury: weather that closes an airport is recorded as an air traffic control delay, not a weather delay, so the true weather contribution sits somewhere inside the 28.3% air traffic control share as well. Even read generously, the airline-side causes remain the larger half. Which raises the obvious comparison.

Is JFK worse than LaGuardia or Newark?

No. LaGuardia is worse than JFK on both measures, and Newark beats JFK on delays. All three run off the same federal dataset and the same twelve months, so this is a like-for-like comparison rather than three separate studies stitched together.

JFK's reputation as New York's problem airport is not supported by its own departure record. The month you travel, however, changes the picture far more than the airport does.

Which month is worst for JFK cancellations?

February 2026 was the worst month for cancellations, at 8.8% of departures, against a twelve-month average of 3.1%. The worst month for delays was July 2025 at 29.1%.

Cancellation risk and delay risk do not move together. Cancellations cluster in winter; delays peak in summer. If you are choosing between them, they are different problems in different seasons, and they also fall differently across the airlines.

Which airline is late most often at JFK?

JetBlue has the highest delay rate at JFK, at 26.4% of its 35,277 departures, against American at 17.6%. These are the airlines' own figures, reported by them to the US Department of Transportation.

Read that alongside flight volume rather than on its own: the busiest carrier at an airport absorbs the most knock-on delay, which is the largest single cause in the table above. Day of week moves the number too, though far less than the hour does: Sun is the worst day at 25.6% and Wed the best at 18.7%.

What this data does not cover

This index covers US domestic departures only, which is roughly half of JFK. Reporting to the Bureau of Transportation Statistics is required only of US carriers above a revenue threshold, and at JFK that is six airlines: JetBlue, Frontier, Alaska, Delta, Republic, American. Every international carrier is absent, and JFK is an international airport.

We are stating that plainly because no equivalent public dataset exists for the international carriers. Anyone publishing a punctuality figure for Emirates or British Airways at JFK is either using a commercial feed they cannot show you or estimating. We would rather name the gap than paper over it, the same way we decline to publish security queue times we cannot verify.

How the JFK Delay Index is built

The index is computed from the US Bureau of Transportation Statistics On-Time Reporting Carrier On-Time Performance dataset, which records every reported domestic flight including the five delay causes. It is US federal public-domain data, which is why we can archive and republish it. Delay is BTS's own 15-minute flag rather than a threshold of our own, so these figures are directly comparable with any other analysis built on the same source.

Every number on this page is generated from the dataset at build time rather than typed in, so the prose cannot drift from the data when the next edition regenerates. Our live figures elsewhere on the site come from a different feed whose licence forbids archiving, which is why the historical index is built on federal data instead. The methods and feeds behind everything we publish are on where our data comes from. The arriving half of the airport, measured the same way from CBP's own reporting, is on the passport gap in the JFK customs queue.

Reusing this? The underlying BTS data is public domain and our analysis is free to quote with a link to this page. The full dataset behind every figure here is downloadable as jfk-delay-index.json, generated by the same script that builds this page, so you can check the working rather than take our word for it. For the live picture rather than the twelve-month one, see current delays and weather at JFK, and to check a specific flight, today’s JFK departures.

Source: US Department of Transportation, Bureau of Transportation Statistics, On-Time Reporting Carrier On-Time Performance, June 2025 to May 2026, accessed 16 August 2026. Public domain.