Urumqi already built: The transit risk for Central Asian airports

Image: Pixabay

China has opened an airport with capacity for 48 million passengers near its border with Central Asia. In 2025, Urumqi handled 29 million people — 61% of its new capacity, almost all of them domestic passengers. Meanwhile, Kazakhstan and Uzbekistan are expanding airports, building hubs and ordering more than $15 billion worth of aircraft. The bet is on transfer traffic. The passengers are still nowhere in sight.

In April 2025, a new airport complex opened in Urumqi, with three runways and annual capacity for 48 million passengers and 550,000 tonnes of cargo. It is the main airport in western China and the nearest major Chinese hub to Almaty and Tashkent.

Urumqi handled 29.18 million passengers in 2025. About one million of them, or 3.4%, were international passengers. Its terminals still have room for almost 19 million more. That is more than Almaty handles in a full year, and roughly as many as Tashkent and Astana handle together. Passenger traffic grew by 4.4% in the first half of 2026.

Across the border, construction is under way too. Almaty is raising $670 million for modernisation and expects to increase annual capacity to 55 million passengers by 2050. Astana plans to build a new airport and has signed an agreement with Changi Airports International. The stated aim is to create an international aviation hub and develop the capital’s transit potential. In Tashkent, Japan’s Sojitz and Saudi Arabia’s Vision Invest are involved in building an airport for 20 million passengers. Private airline SCAT is turning Shymkent into a base for fifth-freedom flights. Since December 2025, SCAT has also been flying from Sanya to Prague under seventh-freedom rights, without entering Kazakhstan at all.

Aircraft have been ordered too. Air Astana could take up to 18 Boeing 787-9s and has signed a contract for 50 aircraft from the A320neo family, half of them options. Together, that amounts to more than $15 billion.

One word is used to justify the airports, hubs and widebody fleet: transit. The question is whose.

Chinese passengers already have their routes

At first, it seems obvious: Urumqi is built, China will keep international traffic within its own network, and Central Asian hubs will have little room to grow. In August, I wrote about how China is moving the boundary of its route network westward. The network is expanding through the major hubs of Beijing, Shanghai, Guangzhou and Xi’an, and through western airports such as Urumqi, Kashgar and Yining (Kulja). For transit through Central Asia, the conclusion is simple: Chinese passengers already have routes that do not need Almaty or Tashkent.

Chinese airlines fly to Europe over Russia, along the shortest routes, which their European competitors cannot use. A connection in Central Asia usually offers those passengers little advantage. They also fly directly to the Caucasus: since July, China Eastern has operated flights from Shanghai to Tbilisi, bypassing Almaty, Tashkent and Urumqi alike.

Local airlines are holding their ground in the direct market. Over the year, Air Astana increased capacity on its China routes by 63%, and FlyArystan by 163%. The market is now almost balanced. But these seats serve direct demand between Kazakhstan and China. In August, I separated two questions: who draws the route map, and who flies it. For hubs, there is a third one: where do passengers change planes?

On China Southern flights, we know the answer. In summer 2024, according to an airline representative, almost 58% of passengers on its flights between Almaty/Astana and Urumqi continued onward within China. Urumqi’s main international destinations are Almaty, Tashkent, Tbilisi, Moscow and Astana. Between January and 21 August 2026, 193,000 passengers were carried on these routes.

So Urumqi already acts as a connecting hub for Central Asia, with room for almost 19 million more passengers. Chinese transfer traffic is not leaving Almaty. Most of it never came through Almaty in the first place.

Three passengers out of a hundred

So what kind of transit traffic are Central Asian airports expecting to grow on?

Official statistics help create the impression of a transit boom. In May 2026, Kazakhstan’s government reported that the number of transit flights had quadrupled over five years and that these flights generated most of the country’s air navigation revenue. It is an impressive figure, but it counts aircraft. Some fly over the country without landing. Others stop to refuel: in late March, jet fuel cost $734–777 per tonne in Kazakhstan, compared with $1,600–1,800 on European and Asian spot markets. Air navigation services, fuel suppliers and ground handlers earn money from these flights. Passengers stay on board. The Urumqi–Zurich cargo service works in much the same way, with a technical stop in Tashkent.

Transit from the south looks similar. On 10 October, VietJet launched Hanoi–Almaty–Prague flights, twice a week on an A330, using fifth-freedom rights. On the return journey, passengers travel on the same aircraft. For them, Almaty is just an intermediate stop. The airport collects fees, while Kazakh airlines face another competitor for passengers travelling to Europe. AirAsia X considered a stop for its European flights somewhere in Central Asia, the Gulf or Turkey, and chose Bahrain. It is too early to call this a trend. We will see how many of these flights remain a year from now.

For now, whichever way you look at it, the region handles other airlines’ aircraft more often than it connects their passengers.

The passengers for whom transfer areas are being built are a different matter. According to IATA data for 2023, out of every 100 passengers arriving in Kazakhstan on an international flight, only three continued to a third country. Six connected to domestic flights — twice as many.

Air Astana has increased its transfer traffic since then. In the first quarter of 2026, the group carried around 48,000 passengers connecting from one international flight to another. Almaty Airport handles about the same number in a day and a half. And the group has to compete with its neighbours for those transfer passengers.

Our passengers change planes in Baku

In the first half of 2026, Azerbaijan Airlines (AZAL) carried 59,200 passengers from Central Asia to Europe via Baku, up 129% from a year earlier. For Almaty, Astana and Tashkent, these are their own departing passengers. For Baku, this is transfer business, and AZAL is building its network around it. Central Asia is designing transfer areas. Baku is already selling connections to Central Asian travellers.

Air Astana’s transfer traffic is growing too. International transfer traffic rose by 65% in the first quarter; in March, after the situation in the Middle East worsened, it rose by 158%. The airline quickly shifted capacity to routes affected by the disruption. The group ended the quarter with a loss of $21.1 million and unit costs up 19.8%. I discussed this paradox in May. The loss cannot simply be treated as the cost of transfer traffic: problems with aircraft availability also affected the result. In the second quarter, transfer traffic grew 82%, the group’s passenger numbers fell 1.7%, and it finished the quarter close to break-even. In its half-year report, the company itself links the growth to changes in its network caused by conflicts in the Gulf and describes itself as a countercyclical airline. A crisis can bring transfer passengers. It can also take them away.

For an airline, this is normal business: an aircraft can be moved to another market within a season. A terminal stays where it is.

The passengers are there. They are local

The airports still need to be built because local passenger demand is growing. Traffic between Kazakhstan and Uzbekistan has risen by more than 30% since the start of the year. Russian airlines announced 67 routes to Uzbekistan for the summer. Migrant workers, tourists and people travelling between neighbouring countries will fill both the new Tashkent airport and the expanded Almaty airport.

In Tashkent, the need for a new airport is clear. It is a necessity, quite apart from the fashion for hubs. The existing airport is surrounded by the city and has no room to expand. It handles almost 10 million passengers a year, and traffic grew 17% in the first half of 2026. There are plenty of questions about the choice of site, but the airport will have to be built.

The question is what is being built beyond that. Astana handled 9.2 million passengers in 2025, more than the design capacity of its current airport. Traffic grew about 2% between January and July 2026. The transport minister plans to start construction in 2027. Yet the purpose has already been set: a hub and transit traffic.

Up to eighteen 787s are another bet on the hub model. Some will replace existing widebody aircraft, but a fleet like this needs transfer passengers to fill it. Direct demand from Kazakhstan can support a few long-haul routes. The rest will have to come from connecting passengers. The first 787 is due in December, while the US service has been delayed again, this time until the second half of 2027. Slots, approvals and route economics still have to be worked out.

Central Asia has its own passengers, and their numbers are growing. Additional international connections will have to be won from other airlines’ networks.

Urumqi as a warning

Back to Urumqi. It has something Central Asia cannot build, however many terminals it adds: a domestic market in a country of 1.4 billion people, a home carrier the size of China Southern, and a government that allocates traffic rights and slots itself. Almaty still has five and a half times as many international passengers — 5.46 million against 0.99 million. But in 2025, international traffic grew 7% in Almaty and 62% in Urumqi, supported by 28 million domestic passengers. Almaty has 6.5 million domestic passengers.

Central Asia is building four hubs without that support. Much of their recent growth in transfer traffic comes from the closed airspace over Russia and crises in the Middle East. The EU sanctions regime has been extended until 31 July 2027. Air Astana’s aircraft deliveries will continue until 2035. Airports are built to last decades longer.

Urumqi is already built. Central Asia is building four more — with no China behind them.

* Kirill Vlasov, air transport expert
https://www.linkedin.com/in/kvlasov/

The views expressed are not necessarily those of CentralAsia+Aero