BTS On-Time Data · January 2018–April 2026

About

Introduction to Know Your Flights

When you're shopping for a flight, you're shown a price, a departure time, and a carrier. What's rarely shown is the history behind that route: how often it runs late, how time of day affects performance, which carriers do better on that specific leg, and what causes the delays. That information exists in public federal records. If you knew that SFO to JFK on a Wednesday morning was historically far more reliable than an evening departure, would you change your itinerary?

Know Your Flights draws from 70 million domestic departures logged by the Bureau of Transportation Statistics between January 2018 and April 2026, across 395 US airports.

How to use KYF

Search a route. Enter an origin and destination, pick a year range, and hit search. You'll see timeliness breakdowns, seasonal and day-of-week delay patterns, carrier comparisons, and year-over-year trends.

Explore the map. Airports are colored by average departure delay rate and sized by annual flight volume. Click an airport to see its outbound routes. Click a route to open its full history.

Search a route → Explore the map

Data Source

All data comes from the BTS Marketing Carrier On-Time Performance dataset. The data starts in January 2018, when BTS switched to the Marketing Carrier format, which attributes flights to the carrier that sold the ticket rather than the one that operated it. Earlier records use a different schema and are not included in KYF. Data is released with a 3 month lag, and KYF includes data until April, 2026. Depending on usage of this tool, the author will maintain the data within this site to include most recent data releases.

View the dataset →   Download the full archive →

Disclaimers

This site serves to explore historical data, not to predict future flight outcomes. It can't tell you what will happen on your specific travel date.

2020 and 2021 flight patterns are largely impacted by COVID-19. With COVID-19 cutting air traffic, 2020-2021 has significantly less delays.