The Best Time to Book International Flights, by the Data

The "book on Tuesday at 3pm" rule is a piece of folklore that has outlived every airline that allegedly invented it. It came from a 2013 study of fares filed by the Airline Tariff Publishing Company, which used to refresh its tape three times a day. ATPCO now updates fares more or less continuously. The rule is dead. It has been dead for a decade. People keep quoting it anyway.

So when is the actual best time to book international flights? The answer is more useful than the myth, but it's also more complicated. It depends on the region, the season, the cabin, and the route's competitive structure. Here's what the data shows once you strip out the noise.

The booking window curve, in plain numbers

The Airlines Reporting Corporation publishes a quarterly study with Expedia using actual ticketing data, not search data. The 2024 update found that for international long-haul economy, the median cheapest booking window opens around 121 days out and closes around 60 days out. Translated: four to six months ahead is the sweet spot.

Hopper's internal data, which is search-and-buy rather than ticketed-only, lines up. They put international economy at roughly 90 to 150 days out, with the steepest fare increases starting around the 45-day mark. Google Flights Insights, which is a different methodology again (aggregated search), shows the same shape with the curve sliding two to three weeks earlier for premium cabins.

Domestic is different. The same ARC data has domestic economy peaking value-wise at 21 to 60 days out. Book domestic too far ahead and you'll pay a premium because the airline hasn't released the cheaper fare classes yet. Book inside 14 days and you're in last-minute business-traveler pricing territory, where a Newark-to-Chicago seat that was $189 three weeks earlier is now $612.

The point is that "book early" and "book late" are both wrong as standalone advice. There's a window, and the window moves with the trip type.

Why the curve looks the way it does

Airline revenue management releases inventory in buckets, not as one big pile. A 777 from Chicago to Frankfurt might have 24 fare classes loaded, each with its own price and seat allocation. The cheapest buckets (call them T, X, V) are released earliest, but in small numbers. As they sell, the system closes them and the next cheapest opens.

Roughly four months out for a popular international route, you're often the first or second buyer into the deeper-discount buckets. By six weeks out, those buckets are usually closed and you're buying out of M, B, or Y, which can be three to five times the price.

There's a second mechanic. Most international fare sales filed by carriers like Lufthansa, ANA, and Qatar are explicitly aimed at the 90-to-180-day booking window. They want to lock in load factor without giving away the late-booking business traveler. So the sales themselves cluster in that window.

The exception is shoulder season and unsold routes. TAP Portugal will run flash sales 30 to 45 days out on Lisbon-to-Boston shoulder dates because the load factor is below target. Norse Atlantic does the same on JFK to Athens in late October. These break the rule, and that's where flight booking window discipline pays off most.

Routes that actually behave differently

Not every international route follows the four-to-six-month rule. Here's where it bends.

US to Western Europe summer peak: book 5 to 7 months out. The window closes earlier because the routes sell out faster. Air France JFK-CDG for June and July typically prices low around January and starts climbing by mid-March.

US to Japan: 4 to 6 months for spring (cherry blossom demand is brutal), but as little as 2 to 3 months in winter. JAL's JFK-HND economy in January 2025 was widely available at $720 round-trip in late November, 7 weeks out.

US to South America: the curve is flatter. Buying 90 days out and 30 days out often produces similar fares because the leisure-business mix is different. LATAM's MIA-GRU in shoulder season barely moves between days 90 and 45.

Intra-Europe short-haul: this is the inverse case. Ryanair, Wizz, and easyJet start cheap and trend up almost monotonically. Book the moment you commit to the dates. Waiting helps no one.

US to India: book 4 to 5 months out for peak (December, summer), but the second cheapest window is paradoxically 7 to 10 days out, when Emirates and Qatar dump remaining seats. This is a high-variance route. Not a route to gamble unless you can be flexible.

The volatility window nobody talks about

The interesting research published by Cirium in 2023 looked at intraday price changes on long-haul international fares. The headline finding: the average international fare changes price 47 times in the 60 days before departure. Some routes change 80 or more.

That means the cheapest day to book is partly a probability statement, not a fixed point. Even within the optimal four-to-six-month window, there are local minima that last 6 to 30 hours. Catching one of those is what separates the cheapest day to book from the next-cheapest.

This is also why the "set a Google Flights price alert and wait" strategy underperforms. Google Flights polls roughly once per day per tracked route. If the dip lasts 12 hours and falls between polls, you miss it.

What to do with all of this

The condensed version. Book international long-haul 90 to 180 days out, with a strong preference for 120 to 150 for peak summer. Book domestic 21 to 60 days out. Set alerts the moment you commit to dates, not when you start "thinking about" a trip. And if you care about catching the intraday dips rather than the rough seasonal average, use a service that polls faster than once a day.

The simplest way to make all of this happen without managing it yourself is to set a watch and let the system do the patience for you. Flyozo monitors inventory across hundreds of routes and pushes you when the price hits the floor, not when the daily refresh happens to catch it.