Guide
Flying in July: the weakest month of the year, airport by airport
July 2026 ran 71.7% on time against 77.3% for the year. Kahului (OGG) led at 86.4%; Fort Lauderdale (FLL) trailed at 62.8%. Where July holds up.
Nearly three US flights in ten arrived 15 minutes or more late in July 2026. The month ran 71.7% on time, against 77.3% across the full twelve months from September 2025 to August 2026, and no other month in the data came in lower. It also carried the heaviest schedule of the year, which is not a coincidence.
July is one month of one year here, so treat the rankings below as a single summer's evidence. The mechanisms behind them are not new, though, and most of the airports at either end of the table are there for reasons that repeat.
The spread between the best and worst airports
US average 71.7%
55.0%90.0%
Flights departing each airport in July 2026, scored on arrival. The dashed line is the national July figure. Source: US DOT Bureau of Transportation Statistics, September 2025 to August 2026.
At the top, Kahului (OGG) ran 86.4% on time. At the bottom, Fort Lauderdale (FLL) managed 62.8%. The median airport sat at 71.8%, and only 4 of the 60 airports with a page reached 80%.
The bottom of the table is dominated by the Southeast and the Washington–New York corridor. Atlanta in July ran 66.5% on time against 77.9% for its year. Charlotte in July fell to 63.8% from 76.1%, and Nashville in July to 64.2% from 74.8%. For all three, July was the worst month of the twelve. Miami (63.6%) and Fort Lauderdale (62.8%) sat alongside them.
Why the Southeast breaks in July
Two things peak at once. Afternoon thunderstorms build over the Southeast almost daily, and they build over the airports that run the biggest connecting banks in the country. A storm cell over Atlanta or Charlotte at 4pm does not delay one flight. It holds an arrival bank in the air or on the ground at the origin, the connecting departures wait for their passengers and crews, and every aircraft in that bank carries the delay into its evening legs. The summer thunderstorm guide covers the hour-by-hour shape of that failure.
The second peak is demand. July is the high point of leisure travel, aircraft are full, and a cancelled flight has no empty seats behind it to absorb its passengers. Across the year, 40% of all delay minutes were charged to an aircraft arriving late from its previous leg. In a month when each leg starts from a later position than the last, that chain runs longer.
The damage is not spread evenly across the carriers at a hub, and the airline-at-airport figures show it plainly.
US average 71.7%
45.0%100.0%
Each bar is one airline's departures from one airport in July 2026. Source: US DOT Bureau of Transportation Statistics, September 2025 to August 2026.
At Charlotte, PSA Airlines, the regional carrier flying American Eagle routes out of the hub, fell further than American's own mainline flights there. Southwest at Nashville had a July far below its own annual record at the airport. Short regional legs in and out of a storm-bound hub are the first to be held, and the first to be cancelled when the airline needs to thin the schedule.
The Northeast: fewer late flights, more cancelled ones
The corridor failed in a different way. Look at the July pages for JFK, LaGuardia and Washington National and the cancellation rates stand out more than the on-time ones: each ran two to three times its own annual cancellation rate (3.98%, 5.04% and 3.99% respectively over the full year). These are runway-limited, slot-controlled airports. When a line of storms crosses the corridor, air traffic control cuts the arrival rate, and the airlines cancel flights rather than run the whole schedule hours late.
That matters for how you read the July table.
A cancelled flight is not scored as late; it is not scored at all.
An airport that cancels heavily can post a better on-time figure than the passengers standing at its gates experienced.
Where July held up
Hawaii sat at the top, as it does in most months. Kahului in July ran 86.4% on time and Honolulu 83.3%. The islands have no afternoon convection of the mainland kind, and the inter-island schedule is a closed network far from the congested airspace that sends delays across the continent. The single best airline-at-airport line in July was Southwest's inter-island flying at Honolulu.
The next rows are less obvious. Columbus (80.9%) and Indianapolis (80.9%) both ran July at or slightly above their annual averages, 80.4% and 80.6%. Neither is a hub, so neither has a bank of its own to jam. Their July cancellation rates were noticeably above their annual ones, though, which flatters the on-time figure a little. Salt Lake City is the strongest hub in July at 79.3%, even though July was also its own weakest month; a bad month in Salt Lake City is still better than an ordinary one at most hubs.
Airlines in July
US average 71.7%
55.0%85.0%
Airlines with no July 2026 flights under their own code are omitted. Source: US DOT Bureau of Transportation Statistics, September 2025 to August 2026.
Envoy Air led at 81.2%, a figure above Envoy's own annual average. JetBlue Airways finished last at 62.2%. Most of the big carriers had their worst month here: Delta's worst month was July at 74.1%, United's August at 73.4%, and American's July at 66.9%.
The Envoy result deserves a note. At O'Hare Envoy's regional flights ran well ahead of American's mainline departures from the same airport in July, a gap the airline-at-airport chart earlier in this guide shows directly. It is not a scheduling artefact. Across the full year, Envoy's O'Hare departures beat American's in every one of the six departure windows, morning and evening alike.
Two airlines are missing from the chart, for different reasons. Hawaiian is reported under its own code only from August to December 2025; from January 2026 the same flights, inter-island included, appear under Alaska's code. I noticed it when the Honolulu to Kahului route split into two carrier files in my pipeline. Spirit's filings stop in early May 2026, so the data has no Spirit flights in June or July. Their annual pages still cover the months they flew.
The Fourth of July week
The holiday itself was not the problem. The Fourth of July week in 2026 ran 72.8% on time, 0.9 points better than the normal weeks either side at 71.9%. Cancellations were a different story: 2.68% inside the window against 1.99% around it. The Fourth of July guide takes the week day by day.
If it were my July ticket
Fly the first wave. In July the morning is the only part of the day the storms have not reached, and it is the only departure window I would book for a July connection. Nationally the 5am–8am window runs 89.3% on time across the year, and in July the gap to the evening is wider, not narrower.
Choose the connecting hub by its July, not its year. Connecting through Minneapolis (77.7%) or Salt Lake City (79.3%) in July 2026 beat Charlotte (63.8%) or Atlanta (66.5%) by a wide margin. Where the fare is close, that choice matters more than the carrier. At 63.8% I would not connect through Charlotte in July if Minneapolis or Salt Lake City were on the table.
Lengthen the connection. The share of flights arriving an hour or more late climbs well above the annual 8.3% in July. The connection risk tool shows how often a given buffer fails on your inbound route.
In the Northeast, plan for a cancellation rather than a delay. Keep both legs on one ticket, book the earliest departure that works, and know the next flight on your route before you leave home.
Hawaii and Salt Lake City are the safe bets. If the trip can go anywhere, the islands and Salt Lake City were the places July did least damage. Denver, Phoenix and Las Vegas were not: all three ran July well below their annual averages.
Check the July pages: Atlanta in July · Charlotte in July · Salt Lake City in July · Honolulu in July · Best time to fly tool