FlyerIntel Research
When to Land to Shorten Your US Customs Wait
Your airline chooses which US arrival hall you land in. You can often choose when, and inside a single hall that is worth more than most travelers would guess.
Published August 30, 2026 · Source period July 2023 – July 2026
Summary
At the 32 arrival halls where enough hours qualify to compare, the gap between the best and worst hour to land is a median of 10.3 minutes at the typical wait and 19.1 minutes on the worst tenth of days. The figures come from 431,971 hourly Customs and Border Protection records covering 61 arrival terminals at 46 US airports between August 2024 and October 2026. At JFK Terminal 1, the most hour-dependent hall measured, landing at 19:00 means a typical 15.6 minutes at passport control and landing at 11:00 means 38.3.
Key findings
- Across 32 major arrival halls, choosing the arrival hour is worth a median of 10.3 minutes at the typical wait and 19.1 minutes on a bad day.
- JFK Terminal 1 is the most hour-dependent hall: 15.6 minutes typical at 19:00 against 38.3 at 11:00, and 31.2 minutes against 61.5 on the worst tenth of days.
- LAX Satellite 7 is the most even, with only 3.2 minutes between its best and worst hour. There the hour is not worth shopping for.
- The worst hour of the day falls between 09:00 and 16:59 at 17 of the 32 halls, and the best hour falls at 17:00 or later at 21 of them.
- Sunday is the best day of the week to land at 19 of the 32 halls with a full week of measurements. Measured against each hall's own week, a Sunday arrival sits 1.5 and 3.4 minutes below (the first figure at the typical wait, the second on a bad day), while Thursday sits 0.8 and 2.0 minutes above.
- Volume is not the whole story: the busiest hour of the day is the worst hour at only 7 of the 32 halls.
A different question from how long the queue is
FlyerIntel has already measured how long US passport control takes and which halls are slowest. This report asks the question that is actually actionable, because the answer to the first one usually is not: your airline chooses your arrival hall, and you cannot overrule it. What you can often choose, when a route is flown more than once a day or on more than one day of the week, is when you land. Customs and Border Protection publishes its wait figures per terminal, per hour and per day, which makes that choice measurable. These are arrival waits at passport control for international arrivals. They are not TSA departure security queues, and the two must never be compared.
How much the arrival hour is worth at each hall
The best and worst hour to land at each hall
| Airport | Terminal | Best hour | Typical | Worst hour | Typical | Swing | Bad-day swing |
|---|---|---|---|---|---|---|---|
| JFK | Terminal 1 | 19:00 | 15.6 min | 11:00 | 38.3 min | 22.7 min | 31.9 min |
| MIA | South Terminal | 23:00 | 8.3 min | 15:00 | 26.7 min | 18.4 min | 33.4 min |
| JFK | Terminal 8 (American) | 18:00 | 14.2 min | 06:00 | 31.4 min | 17.2 min | 22.0 min |
| IAD | International A | 21:00 | 3.5 min | 15:00 | 18.6 min | 15.1 min | 18.8 min |
| JFK | Terminal 4 (IAT) | 22:00 | 10.4 min | 05:00 | 24.8 min | 14.4 min | 28.1 min |
| CLT | Main | 15:00 | 6.6 min | 12:00 | 20.8 min | 14.2 min | 24.2 min |
| HNL | Main Overseas Terminal | 05:00 | 8.7 min | 10:00 | 22.7 min | 14.0 min | 21.6 min |
| FLL | Terminal 4 | 15:00 | 10.3 min | 18:00 | 22.7 min | 12.5 min | 20.1 min |
| IAH | IAB | 21:00 | 10.6 min | 10:00 | 23.1 min | 12.4 min | 22.1 min |
| ORD | Terminal 5 | 22:00 | 11.3 min | 15:00 | 23.5 min | 12.2 min | 26.5 min |
| SFO | Terminal A | 23:00 | 8.7 min | 14:00 | 20.6 min | 11.9 min | 25.3 min |
| MIA | North Terminal | 21:00 | 12.2 min | 10:00 | 23.9 min | 11.7 min | 15.1 min |
| JFK | Terminal 7 (British) | 21:00 | 16.8 min | 16:00 | 28.1 min | 11.3 min | 15.0 min |
| DFW | Terminal D | 21:00 | 8.7 min | 09:00 | 19.7 min | 11.0 min | 17.9 min |
| JFK | Terminal 5 (Jet Blue) | 17:00 | 7.6 min | 05:00 | 18.4 min | 10.9 min | 19.9 min |
| LAX | Tom Bradley International Terminal | 17:00 | 16.2 min | 14:00 | 26.6 min | 10.4 min | 22.2 min |
| BOS | Terminal E | 23:00 | 11.8 min | 09:00 | 22.1 min | 10.3 min | 20.9 min |
| EWR | Terminal B | 17:00 | 12.6 min | 21:00 | 22.8 min | 10.2 min | 22.0 min |
| MCO | Terminal C | 22:00 | 27.0 min | 18:00 | 36.6 min | 9.6 min | 19.8 min |
| LAS | Terminal 3 | 12:00 | 12.3 min | 20:00 | 21.1 min | 8.7 min | 19.4 min |
| ATL | Concourse F | 11:00 | 4.6 min | 20:00 | 12.1 min | 7.5 min | 12.9 min |
| SEA | South Satellite | 20:00 | 16.7 min | 08:00 | 23.5 min | 6.8 min | 9.0 min |
| PHX | Main | 16:00 | 2.3 min | 18:00 | 9.1 min | 6.8 min | 12.4 min |
| SFO | Terminal G | 18:00 | 10.2 min | 07:00 | 17.0 min | 6.8 min | 14.1 min |
| DTW | McNamara Terminal | 16:00 | 9.0 min | 10:00 | 14.5 min | 5.5 min | 11.0 min |
| DEN | International | 19:00 | 8.3 min | 11:00 | 13.8 min | 5.5 min | 9.5 min |
| MCO | Airside 4 | 12:00 | 11.3 min | 19:00 | 16.3 min | 5.0 min | 12.5 min |
| EWR | Terminal C | 20:00 | 10.4 min | 13:00 | 15.2 min | 4.8 min | 11.6 min |
| PHL | Terminal A | 14:00 | 5.7 min | 16:00 | 10.4 min | 4.7 min | 14.9 min |
| MSP | Terminal 1 Lindbergh | 20:00 | 9.7 min | 17:00 | 14.0 min | 4.3 min | 9.4 min |
| ATL | Concourse E | 15:00 | 5.1 min | 18:00 | 8.4 min | 3.3 min | 5.4 min |
| LAX | Satellite 7 | 15:00 | 10.2 min | 18:00 | 13.4 min | 3.2 min | 9.2 min |
An hour qualifies where the terminal has at least 300 measured hours in it and an average of 250 arriving passengers, and a hall appears only where at least 6 hours clear both bars. Bad-day swing is the same comparison at the 90th percentile.
A day at JFK Terminal 1
The shape of an arrival day
The middle of the day is the expensive part. At 17 of the 32 halls measured, the single worst hour to land falls between 09:00 and 16:59, and at 21 of them the single best hour falls at 17:00 or later. It is also when most people land: at 27 of the 32 halls the busiest hour of the day sits between 12:00 and 19:00, as the transatlantic morning wave, the transpacific afternoon wave and the domestic feed overlap. The two facts sit together, but the section below shows that volume accounts for less of this than it looks. The tail moves further than the middle: across these halls the median gap between the best and worst hour is 10.3 minutes at the typical wait but 19.1 minutes on the worst tenth of days, which is the figure that matters if you have a connection to make.
And the day of the week
| Day | Typical wait | Bad day | Against the hall’s own week |
|---|---|---|---|
| Monday | 15.3 min | 28.1 min | -0.3 min |
| Tuesday | 15.2 min | 27.5 min | -0.4 min |
| Wednesday | 15.7 min | 29.0 min | +0.0 min |
| Thursday | 16.5 min | 30.7 min | +0.8 min |
| Friday | 16.3 min | 30.4 min | +0.6 min |
| Saturday | 16.4 min | 30.0 min | +0.7 min |
| Sunday | 14.2 min | 25.3 min | -1.5 min |
Averaged across the 32 halls with a full week of qualifying hours, using each hall's own qualifying hours only, so a hall that operates mainly at weekends cannot tilt a weekday. The last column is the gap between that day and the same hall's weekly mean.
What this is not
These are associations across two years of hourly records, not a mechanism. The obvious explanation for a bad hour is that it is a busy hour, and the data supports that only weakly: measured against each hall's own average, the correlation between how many passengers arrive in an hour and how long they wait is 0.31 across 427 terminal-hours. The busiest hour of the day is also the worst hour at only 7 of the 32 halls. Staffing, the mix of passports on the aircraft that happen to land together, the share of passengers eligible for automated processing, and how tightly the arrivals cluster inside the hour all bear on the result and none of them are separable here. The quiet hours are also quiet because fewer aircraft land in them, and you can only land when your airline flies: an hour that is empty at your hall may simply not be for sale. And a hall's pattern is a property of the schedule as it stood over the observed period: a new route or a retimed bank moves it.
What to do with this
Take the hall as given and shop the hour. Where a route is flown more than once a day, or where a connection can be built onto an earlier or later inbound, the table above says what the choice is worth: a median of 10.3 minutes across these 32 halls on an ordinary day, and 19.1 minutes on a bad one. Budget against the bad-day column rather than the typical one if you have a connection after the queue, because passport control is only the first of several steps between the aircraft door and your next gate. And if the day is negotiable as well as the hour, Sunday arrivals are consistently the lighter ones across these halls.
Pages for what this report covers
Methodology
Calculated by FlyerIntel from 431,971 hourly records published by US Customs and Border Protection, which reports for every arrival terminal and every hour the average and maximum time passengers spent at passport control together with booth counts, flight counts and passenger counts. Those records are first reduced to 5,717 typical conditions (one median and one 90th percentile for each terminal-weekday-hour combination, computed only where CBP published at least eight comparable hours), covering August 2024 to October 2026. Medians are used rather than means throughout, because a single diverted widebody puts a two-hour wait into one hour’s average and misrepresents every ordinary week around it. Hours are combined into a terminal's profile weighted by how many hours were observed, and terminals are never averaged together into a national figure: every comparison in this report is made inside one hall, against that hall's own week. An hour is only compared where its terminal recorded at least 300 measured hours in it and an average of 250 arriving passengers, so that a slot nobody lands in cannot be named as the quickest; a hall is only shown where at least 6 hours clear both bars. Hours are the local hour in which CBP recorded the wait.
Limitations
These are passport control waits for arriving international passengers, not TSA departure security waits, and the two must not be compared. The hour attached to a figure is the hour in which CBP recorded the queue, not the moment an aircraft touched down, so a landing near the top of an hour is described partly by the next one. Figures are typical conditions over the two-year window and mix summer and winter weeks with holiday and ordinary ones; a hall’s pattern reflects the flight schedule as it stood over that period and moves when the schedule does. The measurement covers the passport control queue only, and excludes baggage reclaim, agricultural inspection, re-clearing security and the walk to a connecting gate, all of which sit between an arriving passenger and their next flight. CBP publishes about two days in arrears, so this describes what is normal rather than what is happening now, and nothing here predicts the queue on any individual day. One figure here needs a specific caveat: the wait CBP reports for a terminal-hour blends the US-passport queue and every other queue into one number, and a hall’s own profile page on FlyerIntel publishes those two queues separately, because they can differ by tens of minutes at the same hour, a gap this report’s hour-to-hour comparison does not resolve. What is measured here, and holds regardless, is how much a hall’s combined congestion moves across the day; which of the two queues drives that movement, and by how much, is the profile page’s question, not this report’s.
Sources
- CBP Airport Wait Times — US Customs and Border Protection · source
- FlyerIntel analysis — FlyerIntel · source
Full FlyerIntel methodology · Report a problem with this analysis