FlyerIntel Research
How Long a Layover Needs to Be
The time to reach the next gate is fixed. The margin against a late inbound flight depends on the airport, the route and the hour the flight lands.
Published September 26, 2026 · Source period July 2023 – July 2026
Summary
At ATL, a layover of 74 minutes for a flight due 5–9 AM, or 133 for one due 6–9 PM, keeps the chance of a missed connection to 1 in 20. That assumes 30 minutes are needed after landing and counts only the inbound flight. Across the 20 airports with the most inbound flights, the evening figure is longer at every one, by 43 to 117 minutes. On the 481 routes into those airports with at least 1,000 flights in both windows, the median layover is 66 minutes in the morning and 144 in the evening.
Key findings
- Across the 481 routes into the 20 busiest airports with at least 1,000 flights in both windows, the median layover for a 1 in 20 chance of missing is 66 minutes for a flight due 5–9 AM and 144 minutes for one due 6–9 PM, assuming 30 minutes are needed after landing.
- The evening needs the longer layover at all 20 airports. The difference runs from 43 minutes at SEA to 117 minutes at EWR.
- For flights due 6–9 PM, the layover runs from 102 minutes at SLC to 198 minutes at EWR; for 5–9 AM it runs from 51 to 107.
- 464 of the 481 routes need a longer layover in the evening than in the morning. The median difference is 71 minutes.
- IAD → EWR has the largest difference: 49 min for a flight due 5–9 AM against 269 min for one due 6–9 PM.
- A 60-minute layover carries a 7.0% chance of a miss when the inbound flight is due 5–9 AM and 20% when it is due 6–9 PM, pooled across the 20 airports.
What the number means
Every figure in the tables below is a layover in minutes, the gap between an inbound flight’s scheduled arrival and the next flight’s scheduled departure. It has two parts. The first is the time a traveler needs after landing, to leave the aircraft, walk to the next gate and get there before the boarding cutoff. This report assumes 30 minutes, the default on the connections tool. It is an assumption, not a measurement: FlyerIntel has no gate or terminal data, and a terminal change or a large airport can take longer. The second part is the margin against a late arrival: the delay that no more than 1 in 20 of the route’s inbound flights has exceeded, read from how late its flights have actually landed. The layover is the sum of the two. Only the inbound leg is measured. The data does not know whether the next flight leaves late, whether the airline holds it, or whether the two flights are on one ticket.
Layover for a 1 in 20 chance of missing, into the 20 busiest airports
| Airport | City | Inbound flights | 5–9 AM | 9 AM–12 PM | 12–3 PM | 3–6 PM | 6–9 PM | 9 PM–5 AM | Any time |
|---|---|---|---|---|---|---|---|---|---|
| EWR | Newark | 374,130 | 81 min | 74 min | 105 min | 148 min | 198 min | 175 min | 143 min |
| DCA | Washington | 420,392 | 85 min | 84 min | 113 min | 133 min | 170 min | 153 min | 132 min |
| JFK | New York | 340,949 | 103 min | 74 min | 107 min | 127 min | 168 min | 168 min | 134 min |
| ORD | Chicago | 929,643 | 99 min | 101 min | 116 min | 154 min | 168 min | 161 min | 137 min |
| BOS | Boston | 430,535 | 91 min | 75 min | 103 min | 137 min | 166 min | 155 min | 133 min |
| DFW | Dallas–Fort Worth | 934,529 | 107 min | 107 min | 126 min | 151 min | 164 min | 156 min | 140 min |
| LGA | New York | 438,914 | 66 min | 79 min | 102 min | 138 min | 164 min | 157 min | 130 min |
| MCO | Orlando | 487,683 | 76 min | 78 min | 108 min | 140 min | 161 min | 160 min | 132 min |
| CLT | Charlotte | 605,907 | 79 min | 90 min | 98 min | 127 min | 158 min | 154 min | 124 min |
| SFO | San Francisco | 430,510 | 71 min | 105 min | 114 min | 128 min | 151 min | 142 min | 126 min |
| DEN | Denver | 949,569 | 67 min | 81 min | 87 min | 133 min | 149 min | 132 min | 115 min |
| IAH | Houston | 352,156 | 72 min | 84 min | 97 min | 128 min | 149 min | 145 min | 119 min |
| DTW | Detroit | 382,807 | 84 min | 81 min | 102 min | 117 min | 139 min | 150 min | 118 min |
| ATL | Atlanta | 995,672 | 74 min | 79 min | 87 min | 113 min | 133 min | 144 min | 108 min |
| LAS | Las Vegas | 563,698 | 51 min | 75 min | 101 min | 121 min | 127 min | 130 min | 108 min |
| PHX | Phoenix | 592,677 | 58 min | 76 min | 93 min | 106 min | 125 min | 125 min | 103 min |
| MSP | Minneapolis | 365,449 | 80 min | 74 min | 95 min | 115 min | 124 min | 139 min | 107 min |
| LAX | Los Angeles | 588,685 | 61 min | 74 min | 90 min | 103 min | 122 min | 126 min | 102 min |
| SEA | Seattle | 503,801 | 63 min | 76 min | 83 min | 89 min | 106 min | 114 min | 92 min |
| SLC | Salt Lake City | 352,218 | 54 min | 67 min | 77 min | 91 min | 102 min | 127 min | 91 min |
Minutes, assuming 30 minutes are needed after landing. Columns are the scheduled arrival window in the arrival airport’s local time. Every airport figure pools all of the airport’s inbound routes, so it describes a flight picked at random from those arriving, not a typical route. The airports are the 20 covered airports with the most inbound flights that completed in the period, ordered here by the evening figure.
Extra layover needed for a flight due 6–9 PM compared with 5–9 AM
Chance of missing by layover, all inbound flights into these airports
| Inbound flight due | 45 min layover | 60 min layover | 90 min layover | 120 min layover | 180 min layover |
|---|---|---|---|---|---|
| Any time | 21% | 14% | 7.8% | 5.0% | 2.4% |
| 5–9 AM | 11% | 7.0% | 4.1% | 2.9% | 1.8% |
| 9 AM–12 PM | 13% | 8.3% | 4.5% | 3.0% | 1.6% |
| 12–3 PM | 17% | 11% | 5.9% | 3.7% | 1.8% |
| 3–6 PM | 23% | 15% | 8.7% | 5.5% | 2.7% |
| 6–9 PM | 29% | 20% | 12% | 7.3% | 3.3% |
| 9 PM–5 AM | 31% | 22% | 12% | 7.4% | 3.0% |
The chance that the inbound flight lands more than the layover less 30 minutes late, among flights that operated. Pooled over the 20 airports, 11,039,924 completed flights in all.
What a shorter layover risks
A layover shorter than the figures above is a choice to accept more risk, and the table shows how much. At 60 minutes the chance is 7.0% for a flight due 5–9 AM, 8.3% for one due 9 AM–12 PM, and 20% for one due 6–9 PM. By 120 minutes the morning figure is 2.9% and the evening figure 7.3%. The same connection is a different bet at a different hour. The gap is the accumulation of delay through the day that the site’s departure-time reports measure, seen here on arrivals.
The busiest inbound routes
| Route | Inbound flights | 5–9 AM | 6–9 PM | Any time | Extra in the evening |
|---|---|---|---|---|---|
| SFO → LAX | 34,548 | 54 min | 120 min | 108 min | +66 min |
| LAX → SFO | 34,543 | 54 min | 145 min | 121 min | +91 min |
| LGA → ORD | 32,683 | 66 min | 207 min | 157 min | +141 min |
| LAX → JFK | 30,566 | 101 min | 156 min | 130 min | +55 min |
| PHX → DEN | 28,137 | 49 min | 131 min | 111 min | +82 min |
| DEN → PHX | 28,107 | 62 min | 143 min | 113 min | +81 min |
| LAS → LAX | 28,041 | 55 min | 129 min | 104 min | +74 min |
| LAX → LAS | 28,032 | 50 min | 121 min | 114 min | +71 min |
| DCA → BOS | 27,701 | 63 min | 194 min | 141 min | +131 min |
| BOS → DCA | 27,678 | 62 min | 186 min | 129 min | +124 min |
| MCO → ATL | 26,347 | 56 min | 200 min | 149 min | +144 min |
| ATL → MCO | 26,290 | 71 min | 178 min | 142 min | +107 min |
| DEN → SLC | 24,170 | 60 min | 142 min | 109 min | +82 min |
| DEN → LAS | 24,167 | 55 min | 148 min | 119 min | +93 min |
| SLC → DEN | 24,108 | 71 min | 143 min | 110 min | +72 min |
Layover in minutes for a 1 in 20 chance of missing, assuming 30 minutes needed. The 15 routes into these airports with the most completed flights, among routes that have at least 1,000 in both windows.
Where the evening costs the most
| Route | Inbound flights | 5–9 AM | 6–9 PM | Extra in the evening |
|---|---|---|---|---|
| IAD → EWR | 4,491 | 49 min | 269 min | +220 min |
| MCO → EWR | 19,892 | 58 min | 220 min | +162 min |
| EWR → CLT | 11,272 | 56 min | 214 min | +158 min |
| ATL → EWR | 18,066 | 76 min | 231 min | +155 min |
| TPA → DFW | 8,764 | 64 min | 215 min | +151 min |
| CLE → EWR | 6,060 | 61 min | 211 min | +150 min |
| MCO → JFK | 14,447 | 51 min | 201 min | +150 min |
| MCO → ATL | 26,347 | 56 min | 200 min | +144 min |
| MCO → LGA | 15,455 | 50 min | 193 min | +143 min |
| CLT → EWR | 11,210 | 71 min | 212 min | +141 min |
The ten largest differences among the 481 routes with at least 1,000 completed flights in both windows. Those routes carry 48% of the completed inbound flights into these airports.
What this leaves out
Cancelled and diverted flights are not in the delay record, because they end a connection differently. Among the flights scheduled into these airports, 1.62% of those due 5–9 AM were cancelled or diverted, against 2.30% of those due 6–9 PM. Flights on the same day share weather and air traffic control, so a bad afternoon raises the risk for every connection through that airport at once, and the 1 in 20 figure is an average over days, not a forecast for one. The figures say nothing about the next flight: a departure that leaves late gives a late arrival time to spare, and an airline that holds a plane or rebooks a passenger changes what a miss costs. They also assume the traveler’s own estimate of the time needed. If the walk to the next gate is longer than that, the difference comes off the margin.
Methodology
Calculated by FlyerIntel from 21,708,321 individual flight records in the US Department of Transportation Bureau of Transportation Statistics Reporting Carrier On-Time Performance dataset, covering July 2023 through July 2026. Following BTS methodology, a flight counts as delayed when it arrives 15 or more minutes after its scheduled arrival time. Cancelled and diverted flights have no arrival time and are excluded from on-time rates rather than counted as late; cancellation rates are reported separately. Full definitions are published at /methodology/. The layover is read from the distribution of arrival delay on each route and scheduled-arrival window (FlyerIntel’s route delay table): the count of completed flights that landed at most N minutes late, for every N from -30 to +240. The chance of missing is the share of completed flights that landed later than the layover less the time needed; counts are subtracted before they are divided. The layover for a 1 in 20 chance is the smallest whole-minute layover at which that share is no more than 5%, plus the time needed. This is the same arithmetic the connections tool uses. Windows are the scheduled arrival hour at the arrival airport: 5–9 AM is the morning window and 6–9 PM the evening one. A route-and-window record with fewer than 120 completed flights is left out before anything is pooled (2,036 records holding 178,059 flights, among all covered airports). Airport figures add the counts of every inbound route into the airport, so they are the distribution of the airport’s own arrivals. A route’s own layover is quoted only where both the morning and the evening window hold at least 1,000 completed flights, which puts the 95% interval on the miss chance at the quoted layover at about 3.6% to 6.4%; at 120 flights it would be several times wider. Flights on one day are not independent, so the true interval is wider than that.
Limitations
The source dataset covers scheduled domestic flights operated by carriers above the DOT reporting threshold; smaller carriers and international flights are not included. Flights sold under a mainline brand but flown by a regional carrier are attributed to the operating carrier, which is how the DOT records them. Past performance describes what has happened and is not a prediction about any individual flight. The data has flights, not itineraries: it cannot show whether a passenger actually missed a connection, whether the two flights were on one ticket, or how long the walk between gates was. The time needed after landing is an assumption. Only the inbound leg is measured. Windows are hour ranges, and a flight scheduled across a window edge is counted by its scheduled arrival hour only.
Sources
- FlyerIntel analysis — FlyerIntel · source
- Reporting Carrier On-Time Performance — US Department of Transportation, Bureau of Transportation Statistics · source
Full FlyerIntel methodology · Report a problem with this analysis