International Airfare

A one-way price is not half of a return, a six-stop trip is not six flights, and the cheapest search result is frequently the most expensive way to fly the route (Cash wise and Carbon Wise). International airfare rewards understanding how it is constructed far more than it rewards refreshing a page.

A route is one fare, not a stack of flights

Booked leg by leg, a multi-stop trip is a pile of unrelated one-ways, and one-ways are priced worst of everything airlines sell. Priced as a single itinerary, the same stops go on one fare built across airlines - and the difference between those two numbers on a long route is routinely four figures.

This is the part the big search engines cannot do for you. They price the trip you typed. Nobody types the cheaper version, because finding it means knowing which order the stops go in and which carriers will sit on one ticket together.

Alliance RTW tickets versus a built route

A Star Alliance or oneworld round-the-world ticket is a real product with real rules: a fixed direction, a mileage or continent cap, and every stop inside one alliance's map. When your trip fits that shape it is excellent value. When it doesn't, the rules cost you more than the fare saves - a single stop off the map means either dropping it or buying a separate ticket that undoes the discount.

A route built across airlines has no such map. It is usually the cheaper answer for anything that isn't a clean loop, which is most real trips.

What actually moves the price

The order of the stops. The direction around the planet. How many oceans you cross and how many times. Which two cities you connect by land instead of air. Season, and how far out you buy. Not the day of the week, not clearing your cookies, and not an incognito window.

For years the old version of this page said the opposite. It told people to fly Tuesday through Thursday and to book at least 21 days out, because that was the received wisdom of 2011. It was thin then and it's dead now. What replaced it is below.

Price the gateway, not the destination

Every region has two or three airports carrying real competition and a long tail of airports carrying none. The gap between them has almost nothing to do with distance and almost everything to do with how many airlines want the route.

Nairobi and Kilimanjaro are about 170 miles apart. Fly into Nairobi and you are buying into a market where Kenya Airways, Emirates, Qatar, Turkish and Ethiopian are all fighting over the same seats. Fly into Kilimanjaro and you are buying whatever the single connecting carrier feels like charging that week. The shuttle between them costs less than the fare difference, and it isn't close.

That move works on every continent. Buy the gateway, then get to where you're actually going on a short hop, a budget carrier, or the ground. It costs you an afternoon and it is the single highest-value decision in most international itineraries.

The second layer: flights that exist but won't ticket

If your stops sit on one multi-carrier fare - an alliance round-the-world ticket, or a route built across airlines - something else is going on that no search engine will show you. A flight can operate daily, show seats for sale, and still be unbookable on your fare, because the specific booking classes that fare requires are never open on it.

AirTreks priced this at scale in April 2026 and found 53 legs that fail more than 85% of the time. They are not obscure legs. Several are the most obvious lines on the map, and they are the ones you would type into a search box first. What follows is region by region.

Oceania and the Pacific

Qantas flies Sydney to London and Sydney to New York. On a round-the-world fare you will almost never get either. Sydney to New York came back 0 for 235 attempts. Sydney to London books 6% of the time, Perth to London 10%. Dallas to Sydney was 0 for 74.

These are the flagship routes, which is exactly why people plan around them and then find out at ticketing.

The fix is to stop trying to cross in one hop. Sydney to Los Angeles on Qantas books 80% of the time and Los Angeles to New York on American books 84%. Two legs, both reliable, same trip. Going the other direction, put Singapore or Hong Kong in the middle instead of flying the kangaroo route direct: Singapore Airlines out of Sydney, or Cathay via Hong Kong.

Everything transpacific wants a west coast gateway. Los Angeles, San Francisco or Vancouver, every time.

Asia

Hong Kong is the fix for nearly everything inside Asia. Tokyo looks like the obvious hub if you are coming from North America, and for the Pacific crossing it is. For the onward legs it is not. Japan Airlines from Tokyo to Delhi, Kuala Lumpur, Singapore and Jakarta all price at zero on round-the-world fares. Not low. Zero.

Cathay from Tokyo to Hong Kong books 67%, and Hong Kong connects onward to all of it. Coming from the Gulf, Qatar's Doha to Bangkok leg is the strongest single segment in the entire dataset at 88%.

Hong Kong to New York direct has the same problem in reverse - Cathay on that route was 0 for 162. Route it through Los Angeles.

Air India is the other one to know about. Zero bookability on round-the-world classes, and even when a ticket does get issued their system has a habit of auto-cancelling round-the-world segments that other carriers leave alone. If a flight quietly disappears from your itinerary, that is usually why.

The Middle East

Qatar is an excellent airline and a bad idea as a direct US link on this kind of ticket. Doha to Los Angeles books 10% of the time. Dallas to Doha books 0%. Send the same carrier through Hong Kong, which runs 78%, or through London, and it becomes one of the more reliable airlines on the map.

Separately from any of that: Doha, Dubai and Istanbul are the three gateways that make Africa and South Asia affordable from Europe and North America. That part holds whether you are on one ticket or twelve, and it is the clearest example of the gateway rule anywhere.

Africa

The gateway list for the whole continent is short: Johannesburg, Nairobi, Cairo, Addis Ababa, Casablanca. Everything else is priced off a connection through one of those, so choose the connection on purpose rather than letting a search engine pick it for you.

On an alliance ticket, South African Airways is a trap. Zero bookability. The flights operate, the seats are there, the round-the-world classes never are.

Ethiopian out of Addis Ababa and EgyptAir out of Cairo are the two African carriers that reliably work on Star Alliance fares. Ethiopian's network is the most useful thing on the continent for a multi-stop trip, and it is chronically underused by people planning from outside Africa.

South America

LATAM left oneworld in 2020 and a great deal of published advice never caught up. This is not a partial or occasional thing. Santiago to Los Angeles, to New York, to Miami, to Auckland: round-the-world class availability was zero on all of them. If a plan has you on LATAM as part of a oneworld ticket, the plan is wrong.

American is what actually covers South America for oneworld, and it covers it well - every route probed out of Santiago and Buenos Aires came back healthy in both economy and business. On the Star Alliance side it is Avianca and Copa, which makes Bogotá and Panama City your connecting points.

Central America and the Caribbean

Panama City on Copa is the hub the region hangs off, and Copa is a Star member, so it slots into a round-the-world fare cleanly.

Past that we do not have enough sampling here to tell you much, and would rather say so. In practice most Caribbean routing gets solved with a cheap separate ticket instead of being forced onto the main fare, and that is usually the right answer anyway.

Europe

Europe is the easy one. Competition is thick enough that the gateway trick barely pays and most of the obvious legs book without drama.

Two things worth knowing. Iberia comes in at 10% bookability, low enough that a Madrid connection on an alliance ticket deserves a second look. And once you are on the ground, the cheapest way across Europe is frequently not flying: a surface sector, where you fly into one city and out of another and cover the middle by train, removes a leg from the fare and usually saves money on top.

How many stops before it stops working

The same April 2026 pricing data says a three or four stop route tickets 91% of the time. Five or six drops to 61%. Seven or more falls off a cliff at under 6%.

That is not a rule about what is possible. It is a rule about what an automated system can assemble on its own. Past six stops you are not buying a product any more, you are commissioning one, and it needs a person.

Where to get a route priced

For a straightforward return between two cities, the airfare sites reviewed below are fine and you should just use them. AirTreks exists for the other case: multi-stop routes priced as one fare across airlines. Bring the stops and the rough months.

If you're still working out what the stops should be, the trip planner page starts a step earlier, and the round the world section is the long version.

Bookability figures throughout come from AirTreks' 2026 fare-class probing (updated yearly) across alliance round-the-world inventory. They describe whether the booking classes a round-the-world fare requires are open on a given flight, not whether the flight exists or whether a normal cash ticket is available on it.