---
license: cc-by-4.0
pretty_name: "US Airline On-Time Sample: 10 Hub Airports, One Week, Flight by Flight"
language: [en]
tags: [aviation, flights, on-time-performance, delays, airlines, transportation, tabular]
size_categories: [10K<n<100K]
---

# US Airline On-Time Sample: 10 Hub Airports, One Week, Flight by Flight

Every passenger airline departure from **10 of the busiest US airports**
(ATL, ORD, DFW, DEN, LAX, JFK, SFO, SEA, CLT, MIA) from **2026-09-21 to 2026-09-27**: scheduled departure, actual take-off
and landing, and the delay, for **52,352 flights** by **38 airlines**.
73.6% departed on time.

This is a **fixed sample** published by [PlaneTrack](https://planetrack.ai) so you can check the
format and quality. It is not updated. The full data covers every US airline flight daily since
August 2026 (about 25,000 a day and growing), and is available as a daily feed, an API or a
one-off history, and we can prepare a **free sample for your own airports, routes or airline**:
[planetrack.ai/datasets/](https://planetrack.ai/datasets/#talk).

## Columns

| Column | Meaning |
|---|---|
| `date` | Flight date (FAA schedule date). |
| `flight` | Flight identifier as flown (ICAO callsign, e.g. UAL1234). |
| `airline_icao` | Airline ICAO code. |
| `airline` | Airline name. |
| `airline_group` | us, regional (flies for a major brand) or intl. |
| `origin_icao` | Departure airport, ICAO. |
| `origin_iata` | Departure airport, IATA. |
| `destination_icao` | Arrival airport, ICAO. |
| `destination_iata` | Arrival airport, IATA. |
| `aircraft_type` | Aircraft type code when known (e.g. B738). |
| `scheduled_departure_utc` | Scheduled departure (off-block) time, UTC. |
| `actual_takeoff_utc` | Actual wheels-off time, UTC. |
| `actual_landing_utc` | Actual wheels-on time, UTC, when reported. |
| `departure_delay_min` | Actual wheels-off minus scheduled departure, minutes (includes taxi). |
| `taxi_baseline_min` | Median taxi-out time at the origin over the prior 30 days, minutes. |
| `delay_above_taxi_min` | departure_delay_min minus taxi_baseline_min: the delay beyond a normal taxi. |
| `on_time` | 1 if delay_above_taxi_min <= 15 (the DOT convention), else 0. |
| `delayed_45` | 1 if delay_above_taxi_min >= 45, else 0. |

## How "on time" is measured

FAA SWIM gives the actual **wheels-off** time, not the gate push-back time. Taxiing is therefore
part of `departure_delay_min`. We subtract each origin airport's median taxi-out time
(`taxi_baseline_min`) so only delay beyond a normal taxi counts, then apply the usual 15-minute rule.

## Airlines in this sample (top 10 by flights)

| Airline | Flights | On time |
|---|---|---|
| American Airlines | 10,918 | 70.1% |
| Delta Air Lines | 9,106 | 78.2% |
| United Airlines | 7,982 | 72.7% |
| SkyWest | 5,848 | 75.0% |
| Envoy Air | 3,088 | 80.9% |
| Southwest Airlines | 3,047 | 63.4% |
| Alaska Airlines | 2,450 | 77.6% |
| Frontier Airlines | 1,546 | 69.8% |
| PSA Airlines | 1,238 | 72.6% |
| JetBlue | 1,211 | 73.2% |

## Scope and caveats

- Departures from the 10 airports above, by passenger airlines (mainline, regional, and international
  carriers). Cargo, private and fractional operators are excluded.
- Every row is checked against the FAA LADD list; no aircraft identifiers (hex or registration) are included.
- Cancelled flights are not included (no departure time).
- Times are UTC.

## Source and licence

Derived from FAA System Wide Information Management (SWIM) flight data, processed by PlaneTrack.
Licensed **CC BY 4.0**: use it for anything, including commercially, with credit to
"PlaneTrack (planetrack.ai)".

## Citation

PlaneTrack (2026). *US Airline On-Time Sample: 10 Hub Airports, One Week, Flight by Flight* (2026-09-21 to 2026-09-27). https://planetrack.ai/datasets/