In just a few months, the war with Iran has depleted a large share of the interceptor missile stockpiles the United States spent years building, prompting the Pentagon to launch its most extensive plan in decades to increase air defence missile production.
But the problem cannot be solved simply by building new factories. Missiles consumed within weeks can take years to replace, while production lines face bottlenecks involving rocket motors, raw materials and suppliers. Between increasing production, relying on allies and deploying cheaper forms of interception, Washington is searching for ways to close a gap that may now be better measured in days of combat than in the number of missiles sitting in storage.
Before the war with Iran began in late February, the US military held approximately 452 interceptor missiles for the Terminal High Altitude Area Defence system, or THAAD, according to estimates by the Centre for Strategic and International Studies, CSIS.
After months of intercepting Iranian missiles targeting US bases, CNN reported in August that roughly four-fifths of that stockpile had been consumed, along with about half of the Patriot inventory.
Another estimate came from a US official who told ABC News that the war had consumed two-thirds of the Patriot stockpile and half of the THAAD inventory, describing both as “extremely low”.
The percentages differ, but the conclusion is the same: interceptors are being consumed far faster than factories can replace them. The THAAD production line, for example, currently delivers only about 96 missiles per year.
For that reason, the US Department of Defense announced two agreements on 31 August with Lockheed Martin and General Dynamics’ munitions division, paving the way for seven-year procurement contracts intended to triple Patriot interceptor production and quadruple THAAD production, with a long-term objective of nearly 14,000 interceptors across the two systems.
More money and larger orders, however, do not mean missiles will arrive quickly.
According to CSIS, manufacturing time remains the main constraint, and the first replacement batches of THAAD and Patriot interceptors will not begin arriving before 2029.
US Defense Secretary Pete Hegseth acknowledged the problem, saying rebuilding the stockpiles would take “months and years”, depending on the weapon.
A Gap Measured in Days of Combat
The 2026 war was not the first warning.
During the 12-day war between Israel and Iran in June 2025, the United States fired between 100 and 150 THAAD missiles. That represented approximately 14 per cent of its stockpile according to an estimate by the Jewish Institute for National Security of America (JINSA, and roughly one-quarter, according to CNN.
Each missile costs approximately US$12.7 million, while the United States received just 11 new THAAD missiles during the previous fiscal year.
Before that, the commander of US naval surface forces revealed that American ships had fired more than 200 interceptor missiles in the Red Sea over 15 months, at an estimated cost of about US$1 billion, to counter Houthi missiles and drones, some of which cost as little as US$2,000.
The 2027 budget request then raised the Army’s target PAC-3 inventory from 3,376 missiles to 13,773, roughly quadrupling the planned stockpile in a single step.
But the more useful question is not how many missiles are in storage. It is how many days of combat one full year of production can replace.
When a war consumes in weeks a quantity of missiles that requires years to reproduce, headline stockpile numbers become less reassuring.
This exposes the central dilemma: the interceptor is expended in battle like ammunition, but manufactured as though it were a satellite.
Operational doctrine often assigns two interceptors to a single incoming target. Yet manufacturing relies on manual assembly, testing of individual units, a limited supplier base and lengthy production cycles.
An industrial system designed to produce small numbers of highly complex weapons therefore cannot suddenly be converted into one capable of producing thousands overnight.
A Production Race That Will Take Years
The most obvious solution is to increase production.
The August agreements aim to raise annual PAC-3 output from approximately 650 missiles to about 2,000, while THAAD production would rise from 96 to 400 missiles per year.
The total value of Lockheed Martin’s multi-year Patriot programme has exceeded US$58 billion, alongside a new facility in Arkansas intended to accelerate munitions production.
General Dynamics, meanwhile, is expanding output of key components, including motor cases, guidance housings and centre sections.
The significance of the agreements also lies in their duration.
A seven-year contract gives companies greater “demand certainty”, encouraging them to invest in new factories, equipment and workers instead of relying on contracts renewed annually. It is one of the lessons Western military procurement systems have drawn from the war in Ukraine.
That certainty remains incomplete, however, because the Pentagon has made clear that funding will still depend on congressional approval each year.
The contract may span seven years, but the money arrives one year at a time, leaving investment exposed to Washington’s annual budget cycle.
Rocket Motors: The Hidden Bottleneck
Even if assembly lines expand, a deeper problem remains: rocket motors.
Almost every guided American missile depends on a solid-fuel rocket motor produced by one of only two companies, Northrop Grumman or L3Harris, which acquired Aerojet Rocketdyne.
Breaking Defense has described this duopoly as one of the most serious bottlenecks in the US missile industry. Under traditional production methods, even replacing what has been sent to Ukraine alone could take more than a decade.
Washington is attempting to expand both companies’ production capacity.
Northrop Grumman is seeking to raise output from approximately 13,000 motors in 2024 to 25,000 by 2029, while L3Harris is constructing a new US$400 million complex in Arkansas.
At the same time, the United States is trying to introduce new suppliers.
Anduril has opened a factory in Mississippi that aims to produce 6,000 tactical rocket motors annually by the end of 2026.
X-Bow has received contracts worth more than US$250 million as it seeks to become a third approved supplier, while Ursa Major is pursuing a model under which it manufactures most components internally.
The US government has made eight investments in the sector through the Defence Production Act, totalling approximately US$120 million.
But another bottleneck sits behind the motors themselves.
Ammonium perchlorate, which supplies solid rocket fuel with the oxygen required for combustion, is produced by only one plant in North America, owned by AMPAC in Utah.
The company has announced a US$100 million expansion expected to increase capacity by more than 50 per cent, with completion planned during 2026.
Turning to Allied Factories
Another option is to distribute production beyond the United States.
In November 2025, Japan delivered to Washington the first batch of Patriot missiles manufactured on Japanese territory, replacing missiles the United States had sent to Ukraine.
Mitsubishi Heavy Industries, the only company licensed to manufacture the missiles outside the United States, produces about 30 missiles annually. The two countries are also negotiating joint production of the newer MSE version.
In Europe, the COMLOG facility in Bavaria, Germany, a joint venture between Raytheon and MBDA, is preparing to begin production at the end of 2026 and deliver its first batches in early 2027.
It will become the first Patriot production line in Europe and is intended to fulfil a joint order for 1,000 GEM-T missiles worth US$5.5 billion.
This approach adds manufacturing capacity beyond US factories and spreads the risk created by rapid depletion.
It does not, however, eliminate licensing and export restrictions. In any major war, producing countries may also prioritise their own national requirements over exports.
The Case for Cheaper Interception
Rather than waiting for factories to catch up, there is a faster option: use a cheaper weapon against a cheaper threat.
Over the Red Sea, US fighter aircraft used laser-guided APKWS rockets against drones at a cost of approximately US$35,000 per missile, instead of air-to-air missiles costing between US$500,000 and US$1 million.
According to reports, APKWS accounted for nearly half of the interceptions during Operation “Rough Rider”, while F-15E aircraft carrying 42 of the rockets per sortie are being tested.
In Ukraine, German Gepard anti-aircraft guns shoot down Shahed drones using ammunition costing between US$300 and US$1,000 per round.
Ukrainian commanders also say interceptor drones are responsible for bringing down approximately 70 per cent of Shahed drones.
Directed-energy weapons offer another option. Israel’s Iron Beam laser system has a firing cost that is effectively limited to the electricity required to operate it, although its use during the 2026 war remained limited because only a small number of batteries had been deployed.
Robert Tollast, a researcher at the Royal United Services Institute, summarised the problem when commenting on NATO fighter aircraft being used against cheap Russian drones over Poland, saying the asymmetric cost equation “doesn’t work”.
The advantage of cheaper alternatives is that they can immediately reduce the rate at which expensive missile inventories are consumed.
Their limitations are equally important. Such weapons are primarily suitable for drones and slower cruise missiles. They cannot replace THAAD and other heavy interceptors against ballistic missiles.
The Cheapest Missile May Be the One You Do Not Fire
All of these solutions address the supply side of the problem: how can more interceptors be produced faster and at lower cost?
But another side receives less attention: demand for the interceptors themselves.
If every object launched by an adversary automatically triggers the launch of an interceptor, then the adversary effectively controls the rate at which your stockpile is depleted.
The attacker chooses the timing and scale of the attack, and a relatively inexpensive workshop producing drones and missiles becomes, in effect, an order book for much more expensive defence factories.
Insurance provides a useful analogy.
There are rare but catastrophic risks, such as earthquakes, and lower-cost but recurring risks, such as scratches on a car. The first can justify expensive insurance because the potential loss is enormous. The second must be managed differently because attempting to prevent every minor loss can become more expensive than the loss itself.
A ballistic missile resembles the first category. Using an expensive interceptor against it can therefore be rational.
Large numbers of cheap and repeatedly deployed drones are closer to the second category.
The problem is that much of the West’s heavy air defence architecture was designed to confront catastrophic threats. The threat environment then changed as inexpensive drones and missiles proliferated in large numbers, while the response sometimes continued to rely on the same costly weapons.
Israel’s Iron Dome offers a different model.
The system calculates the trajectory of an incoming rocket and its expected point of impact. If the calculations indicate that it will fall in an uninhabited area and does not represent a threat worth intercepting, it is allowed to land without an interceptor being fired.
The least expensive solution may therefore be to change how existing stockpiles are used.
Sensitive installations and populated areas can receive concentrated protection. Cheaper defensive systems can be assigned to cheaper threats. Incoming objects unlikely to cause serious damage can sometimes be allowed to pass.
THAAD, SM-3 and other heavy interceptors can then be preserved for ballistic missiles and the highest-value threats.
