Could the U.S. and China go to war over Taiwan?


Could the U.S. and China actually go to war over Taiwan? Imagining the unimaginable.

A war over Taiwan would likely involve the largest and most complex amphibious invasion ever mounted. Were the conflict to drag on, it might well evolve into a building-to-building, mountaintop-to-mountaintop ground war in one of the most densely populated and economically advanced countries on Earth. And that’s just in Taiwan itself.

Hear more from Joshua Keating about this story:

It’s an open question whether the U.S. would come to its longtime ally Taiwan’s aid; if the United States got involved, we would see a scenario the world has managed to avoid over the 75 years since the introduction of the atomic bomb: direct exchange of fire between two nuclear-armed superpowers.

“Disabuse yourself of the notion that war with China is going to be like anything we’ve experienced in our lifetimes,” said David Ochmanek, a former deputy assistant secretary of defense in the Obama administration who is now a senior researcher at the RAND Corporation.


China’s desire to retake Taiwan goes back decades

While still far from inevitable, this nightmare scenario has never seemed more likely. Beijing has sought control of Taiwan, which it considers a wayward province, ever since 1949, when fleeing Chinese nationalist forces set up a government on the island.

China has unsuccessfully attempted military force against Taiwan before, in the 1950s and 1990s. For much of that period, Taiwan itself had a superior military to the People’s Republic, and U.S. naval dominance in the region was unquestioned.

But in recent years, the balance of power has shifted dramatically, thanks to China’s economic rise and one of the largest and fastest military buildups in history. Until a few years ago, most experts believed China had little chance of successfully taking Taiwan in the face of U.S. opposition. Now, as Ochmanek has put it, the U.S. regularly “gets its ass handed to it” in Pentagon war games simulating a Chinese invasion of Taiwan.

Lately, Chinese leaders have shown signs that they are running out of patience. Chinese President Xi Jinping has said on more than one occasion that the Taiwan issue should not be “passed on from generation to generation.” More recently, he emphasized: “The historical task of the complete reunification of the motherland must be fulfilled and will definitely be fulfilled.” Xi would undoubtedly prefer to accomplish this without war, but that seems increasingly unlikely.

Public sentiment in Taiwan is turning ever more against Beijing — around two-thirds of the country’s population now identify as “Taiwanese” rather than “Chinese,” according to one recent poll — and in 2016, Tsai Ing-wen, who ran on a platform of defending the nation’s sovereignty, was elected president. And recent crackdowns in Hong Kong have undermined Chinese assurances that Taiwan would be allowed to maintain some degree of autonomy under a “one country, two systems” model.


“If you’re sitting in Beijing, it’s pretty clear that anything short of direct military force is not going to be adequate,” Elbridge Colby, a former deputy assistant secretary of defense in the Trump administration, told Grid. “In fact, the trajectory is moving exactly in the opposite direction.”

Examples of that trajectory are in plain sight. In recent months, China has shown signs it is preparing for the use of force. Starting in October, planes from the People’s Liberation Army (PLA) Air Force made a record number of incursions into the area surrounding Taiwanese airspace. The PLA air force has expanded air bases at Longtian, Huian and Zhangzhou, directly across the strait from Taiwan. Taiwan’s government agencies currently face around 5 million cyberattacks per day, around half of them believed to originate in China. China has not been subtle in its rhetoric or its tactics: Satellite images have shown a scale replica of Taiwan’s presidential office building constructed in Inner Mongolia for PLA soldiers to practice raids.

None of these developments suggests that Xi has made the decision to take Taiwan by force. But Xi has described reunification with Taiwan as a core component of a larger political project called the “great rejuvenation of the Chinese nation” (think, “make China great again”) due to be completed by the People’s Republic’s 100th anniversary in 2049. Time and again, he has publicly tied his own political fortunes to the goal.

The view from Washington

The hard line from Beijing is met regularly by a hard line in Washington. As part of a larger shift toward “great power competition” with China, both the Trump and Biden administrations have elevated Taiwan as an issue.

Technically speaking, the U.S. does not recognize Taiwan as an independent nation. The island is part of China under a long-held formulation known as the “One China” policy, but the U.S. is also bound by law to provide aid for Taiwan’s defense. The U.S. — under Republican and Democratic administrations alike — has long been deliberately coy about what it would do if China invaded, under a policy known as “strategic ambiguity.”

President Joe Biden has, on more than one occasion, been less ambiguous, saying the U.S. has a “commitment” to come to Taiwan’s aid were the island to be attacked. The Biden team invited a delegation from Taiwan to its Summit For Democracy in December, enraging Beijing. Support for Taiwan has lately been a rare example of bipartisan agreement on Capitol Hill, and delegations from both parties have traveled to Taipei in recent months. In the military realm, it was reported in October that U.S. special operations troops and Green Berets had been in Taiwan for over a year, secretly training its forces to resist potential Chinese aggression.

The outgoing U.S. Indo-Pacific commander, Adm. Phil Davidson, suggested in congressional testimony in March 2021 that China is likely to move on Taiwan “in the next six years.” Some experts feel that as U.S. support ratchets up, and because of steps Taiwan and the U.S. are taking to fortify the island’s defenses, China may want to move sooner than that. Whatever the timetable, analysts agree that the risk of miscalculation leading to conflict is great.

Prelude to attack

Analysts and former officials told Grid they envision a range of scenarios for how the first few days of a conflict over Taiwan might play out. War-gaming is always fraught with hypothetical and shifting scenarios, but several common threads emerge.

One thing the Chinese military would probably not have on its side is the element of surprise. Defense analyst Ian Easton, whose 2017 book, “The Chinese Invasion Threat,” imagines what war might look like based on leaked Chinese military documents, suggests that somewhere between 1 to 2 million combat troops would have to cross the strait if Taiwan’s defenses were at full strength. (If China already had Taiwan on the back foot by instigating a coup or assassinating its president, a smaller force might be feasible.) The 1944 Normandy invasion, by contrast, involved a landing force of some 132,000 troops.

A Chinese landing on Taiwan would “be the most complex operation in modern military history,” said Michael Beckley, a professor at Tufts University and fellow at the American Enterprise Institute who studies U.S.-China competition.


China’s navy — the dramatic recent buildup notwithstanding — doesn’t have enough amphibious assault ships to transport even a fraction of that number of troops, so it’s anticipated that civilian ships, including passenger ferries and even fishing boats, would be pressed into service. The logistics alone would be staggering. Food, fuel and medicine would be stockpiled. Some of these preparations could be camouflaged as military drills, but most such movements would be obvious to the outside world, days or weeks in advance.

Before any land invasion, China would attempt to achieve dominance in the air, sea and “electromagnetic” realms. Barrages of rockets, missiles and armed drones would target the island’s air defenses — including air base runways and radar installations — in an attempt to ground the Taiwanese air force. Air and sea offensives would be launched against Taiwan’s navy, and China would almost certainly attempt to impose a naval blockade of the island. Military and political targets on Taiwan would be bombed in the early days of an offensive — likely including major government buildings in Taipei — as would the country’s power grid and fuel supplies. Perhaps most important, bombers would be used to degrade Taiwan’s coastal defenses before an invasion. According to an open-source 2015 estimate, there are roughly 40 air bases from which Chinese fighters could strike Taiwan without refueling. Even with a substantial number held back for reinforcements, as many as 800 aircraft could be dedicated to a Taiwan campaign.

Less visible — but potentially almost as damaging — Taiwan’s civilian and military infrastructure would be the target of massive cyberattacks. Taiwanese analysts have also warned that undersea cables linking Taiwan to the global internet could be cut by the PLA in the lead-up to an attack.

For China, the goal in these early days would be to sow as much chaos as possible to either soften Taiwan for invasion or, better yet from Beijing’s point of view, convince its leaders and people that it’s not worth fighting at all.

What will the U.S. do?

No question preoccupies leaders in Taipei, Beijing and Washington more than this: How will the U.S. respond?


Taiwanese military estimates suggest its forces might hold out on their own against China from two weeks to about a month. Even that may be optimistic. If the U.S. enters the fray, it becomes a much fairer fight. And it raises the most critical strategic questions for Beijing.

If Chinese leaders are certain that the U.S. commitment to defend Taiwan is ironclad, their best move will be to strike first. A 2017 war game conducted by the Center for a New American Security (CNAS) found that a preemptive Chinese missile strike could “crater every runway … at every major U.S. air base in Japan, and destroy more than 200 aircraft on the ground.” The RAND Corporation has estimated that just 36 missiles could shut Kadena Air Base in Okinawa, Japan, to fighter traffic for four days. In 2015, the PLA unveiled an intermediate range ballistic missile that was quickly nicknamed the “Guam killer” for its potential ability to reach Andersen Air Force Base on that island. The RAND study found that if 50 of these missiles struck their targets, it could keep the base closed to large aircraft for eight days.

Of course, one act of aggression under these circumstances would almost certainly beget another — and the Chinese know this. Attacking U.S. forces, particularly forces on U.S. territory in Guam, would all but guarantee an American military response.

On the other hand, if the Chinese believe the Americans are bluffing, and that the U.S. has no real interest in a Pacific war, the better move would be to attack the island without provoking the Americans. If no Americans have been harmed, will a U.S. president really launch a war against a rival superpower, risking American lives and billions of dollars in a fight the U.S. might well lose? Polling shows Americans evenly divided — though the numbers have risen lately — as to whether they would support using U.S. troops to defend Taiwan.

But if the U.S. president does decide that defending a democratic ally and halting China’s military expansion is worth the fight, China will have lost a crucial opportunity to strike an early blow and will face a much tougher fight once the Americans show up. (Worth noting: Unlike the U.S. military, the PLA has virtually no combat experience. China has not fought a major war since an ill-fated invasion of Vietnam in 1979.)


Assuming U.S. forces are not badly damaged in an initial onslaught, the Americans have a range of options for striking China in the early days of a Taiwan invasion. It might be days before they reach the area — depending on advance intelligence and positioning — but U.S. submarines, surface ships, aircraft and cruise missiles would likely sink a significant number of China’s amphibious craft during the crossing of the strait. As of 2017, RAND estimates showed that U.S. submarines alone could destroy almost 40 percent of the Chinese amphibious fleet during a weeklong conflict, though China’s defenses in this area have improved significantly since then. If they were able to penetrate Chinese airspace, U.S. aircraft could attack those 40 Chinese air bases within range of Taiwan, though this would represent a significant escalation. Beyond conventional warfare, the U.S. and China could also carry out attacks on each other’s satellites and cyberattacks in the early days of a conflict.

A U.S. intervention hardly guarantees victory, however. The “blue” teams representing the U.S. have repeatedly lost to Chinese “red” teams in classified Air Force war games held since 2018. This is partly by design — the games are designed to highlight vulnerabilities — but they also highlight real strategic dilemmas. China will enter the conflict with a major geographic advantage: Taiwan lies only a few hundred miles from the full strength of the PLA; it’s more than 1,500 nautical miles from Guam, the closest U.S. territory.

In addition to its geographic advantages, China in 2010 became the first country in the world to announce it had operational anti-ship ballistic missiles, capable of precision strikes on U.S. ships at a range of more than 900 miles. These missiles are part of a suite of systems designed to prevent U.S. forces from entering or maneuvering within an area of combat — a concept known in military jargon as anti-access area-denial or A2/AD.

Recent U.S. defense planning has been focused on penetrating these defenses, but according to a recent Defense News report, the most promising Pentagon scenario for repelling an invasion of Taiwan relies on air defense technology that doesn’t exist yet.

The crossing

The most dangerous part of any invasion from the Chinese perspective would be the actual crossing of the 100-mile strait. “I certainly wouldn’t want to be in one of those Chinese transport crafts, puttering across the Taiwan Strait for eight hours,” Tufts University’s Beckley told Grid. While the Chinese would pound Taiwan’s coastal defenses during the initial air assault, they would be unlikely to eliminate them. Several analysts noted that the vaunted U.S. Air Force had a difficult time eliminating Iraqi and Serbian mobile missile batteries during the Gulf and Kosovo wars; Taiwan’s military is far more advanced.


During this period, the China-Taiwan air war would continue to be intense. “There’s no real way to not be vulnerable during the actual amphibious invasion,” Cristina Garafola, who researches the Chinese military at RAND, told Grid. “You’re basically sitting ducks for anti-ship cruise missiles or Taiwanese aircraft. So I think the [PLA air force] and the PLA navy’s aviation branch would have a key role in making sure those ships made it across to Taiwan with as few casualties and as little damage as possible.”

When China’s ships actually land on Taiwan’s beaches, they’re likely to be met with mines, nets and traps along the shoreline. In one of the more vivid sections of his book, Easton cites PLA military documents that describe Taiwanese plans to create “sea walls of fire” by pumping oil into the shallows and lighting it aflame. (These reports are unconfirmed.)

China’s best bet for moving troops and supplies onto the island will be to seize a port. That means the Taiwanese may have to take the drastic step of destroying their own ports to prevent a landing.

“[The Chinese] have to have that combo of sufficient combat power ashore plus air superiority,” said Chris Dougherty, a senior fellow at CNAS, who has designed numerous Taiwan war simulations. “In almost all of our war-games, there’s a big scramble to see if Taiwan can destroy their ports and their airports fast enough.”

There’s precedent for such operations — the retreating Germans left the port of Cherbourg in ruins to deny the allies a base of operations after D-Day — but the scale and implications of the task should not be downplayed: Taiwan’s largest port, Kaohsiung, is among the 15 largest in the world, handling a larger annual volume of shipping than the Port of Los Angeles.


The land war

Most experts agree that a short, sharp war offers China’s best bet for victory. “If you read Chinese military writings, they talk about essentially the equivalent of ‘shock and awe,’” Beckley said. “I think the Chinese have a reasonable expectation that if they just hit Taiwan and the U.S. hard enough in the face, that they will be paper tigers and stand down.” If China can cripple Taiwan’s resistance quickly, the thinking goes, the international community will be hard pressed to do much about it. Defense scholars refer to this is as a fait accompli strategy, best demonstrated in recent years by Russia’s rapid and almost uncontested annexation of Crimea in 2014.

As Eric Heginbotham, a researcher on Asian security issues at the Massachusetts Institute of Technology, notes, the longer a conflict drags on, the tougher it becomes from Beijing’s perspective: “Once you start lethal action, then Taiwan begins to mobilize its reserves. And as long as that ground campaign is going on, your fleet has to be anchored off the coast where it is vulnerable. Any chance for success really does count on a quick, relatively quick occupation.”

Taiwan has a standing army of around 190,000, but hundreds of thousands more troops could be pressed into service from its reserves. That’s still not many compared with the largest military on the planet, but the Taiwanese would have the advantage of fighting on their turf, in mountainous terrain that could be a nightmare for occupiers.

Polls suggest the island’s population is overwhelmingly willing to take up arms, but that could change quickly if defenses were being overwhelmed or if U.S. support looked iffy. “On our side, the will is there. But it’s that’s not enough. You need it to be trained and prepared ahead of the contingencies,” said Enoch Wu, a former Taiwanese special forces soldier turned politician who has founded an organization advocating greater preparedness for war among average citizens.

Meanwhile, a drawn-out conflict would give Taiwan’s allies — the U.S. but also potentially Japan or Australia — more time to organize support and chip away at China’s air superiority. “Over time, we will have superior aircraft, superior pilots,” Ochmanek said. “I say to people, you can’t win this war in a week, but you can lose it in a week.”

So what happens if — in war-game parlance — the “blue” team stands up, Taiwan mobilizes its reserves, and the U.S. and other allies provide air and naval support? In all likelihood, China still would land a large number of troops on the island, but after that many of the scenarios and planning documents grow vague. It’s been decades since the world has seen mechanized ground combat in urban areas between two heavily industrialized powers, said MIT’s Heginbotham: “Honestly, I think it looks a lot like it always has or has for the last 100 years — which is bloody and slow.”

CNAS’s Dougherty said that war games that get to the ground war stage often settle into a “kind of stalemate” in which the Chinese have control of the air but the Taiwanese are dug in with armor and artillery. “The Chinese can make some headway when they have extra air superiority, when we don’t have fighters up. But then, every 12 to 18 hours, we get a big sortie of bombers, and they get pushed back again. And it kind of goes back and forth like this over the course of days and days and days.”

No refuge for civilians

Less discussed in the war games and white papers is the potential impact on Taiwan’s civilian population. Beyond the fighting itself, the damage could prove catastrophic. Taiwan imports much of its food, fuel and medicine, and an extended blockade could have devastating humanitarian consequences.

In his book, Easton notes that civilian casualties are hardly mentioned in the PLA documents he studied. He told Grid that on a recent trip to Taiwan, he couldn’t help but think about which of the places he was visiting would be targeted in the event of war: “When you do this kind of research, it changes the way you look at the world. The whole country is going to be a war zone.”

With a population of 23.5 million in an area a bit bigger than Maryland, Taiwan is one of the densest countries on Earth and is particularly so along its west coast, where Taiwan’s main military installations are located and where the fighting is likely to be most intense.

“Our military bases are in our cities,” Wu told Grid. “Our communications networks, our infrastructure, our power plants, those are all strategic targets. All those nodes are in urban environments. We all recognize that if conflict breaks out, there are no safe zones.”

The nuclear question

To some of the experts, these grim scenarios are actually good news, in that China is unlikely to embark on a mission that has a good chance of turning into a gruesome stalemate. The problem, given the importance of reunification as a goal for the Chinese state, is the difficulty of getting Beijing to back down.

“How many times does Xi Jinping have to say that the destiny of China is to reunite Taiwan with the mainland before people are convinced that he really means it?” Ochmanek asked. “Once they commit to this course of action, they have a lot riding on it.”

Meanwhile, looming over any discussion of escalation is the fact that both the U.S. and China have nuclear arsenals. (China’s is much smaller but catching up.) In his new book, “The Strategy of Denial,” which argues the U.S. needs to be doing more to deter a Chinese “fait accompli” attack on Taiwan, Colby, the former deputy assistant secretary of defense, argues that any war between the U.S. and China would likely be limited in critical ways, because the incentives to avoid massive nuclear exchange are too great for both sides. He argues this isn’t necessarily a good thing; you might be more likely to start a war if you believe you can keep it under control.

Beckley is skeptical that war would remain limited. He worries about a scenario in which “both sides assume the other side would never go nuclear. And so, it’s OK to just hit them as hard as you want with conventional forces, because there’s a fire break between that the nuclear realm.”

Colby accepts that the risk of nuclear escalation is real but said it shouldn’t deter the U.S. from defending Taiwan. “If we are completely convinced that a limited war is impossible, and the Chinese believe that it is possible, then they will checkmate us every time,” he said. “At some point, we have to be willing to fight a war under the nuclear shadow. My view is the best way to avoid testing that proposition, which I absolutely don’t want to do, is to be visibly prepared for it.”

For some U.S. analysts and policymakers, the risks to Taiwan (and in a nuclear scenario, the risks to everywhere else) are an argument for bolstering support for the island, making an invasion seem as unappealing a prospect as possible to Beijing. For others, the risks involved in defending Taiwan are simply too great, and the U.S. should cut the island loose from its strategic priorities. Either way, it’s best for all involved if these scenarios remain theoretical.


In the end, these scenarios are influenced — in Beijing and Washington both — by politics and long-standing principles. For China, it’s that article of faith about Taiwan, and “One China,” which has recently been married to a position of geopolitical and military strength. For the U.S., it’s the defense of an ally and an increasingly vigilant stand against China and its regional ambitions. Neither side wants war. Neither side has shown any interest in standing down. The war-gamers are still in business.

This article has been updated.

  • Joshua Keating
    Joshua Keating

    Global Security Reporter

    Joshua Keating is a global security reporter for Grid focused on conflict, diplomacy and foreign policy.