Why America Has a Launch on Attack Option


If Russia were to maintain its nuclear forces at current levels, it would take more than its entire land-based intercontinental ballistic missile force to destroy America’s land-based intercontinental ballistic missiles. This requirement serves as an effective deterrent, since it makes an attack on American missile fields a high-risk option. Unfortunately, the New START Treaty that bound Russia to limits on operationally deployed strategic nuclear weapons faltered amidst serious tensions following Russia’s invasion of Ukraine and the subsequent decision to suspend the treaty’s implementation. 

We believe Russia’s continued observance of New START Treaty limits is increasingly unlikely. Russian President Vladimir Putin could rely more on nuclear weapons to compensate for his declining conventional performance in Ukraine. Should Russia do so and, on the worst day, choose to preemptively strike the U.S. nuclear arsenal in a crisis, President Putin has a range of options to employ against America’s intercontinental ballistic missile force. For this and other reasons discussed below, we believe that the United States should keep its intercontinental ballistic force “on alert” and maintain its “launch under attack” option to both ensure the force’s survivability in a conflict and deter adversaries from seriously contemplating a first strike. 

In a March 17 War on the Rocks article, “Launch Under Attack: A Sword of Damocles,” Natalie Montoya and R. Scott Kemp recommended eliminating the U.S. intercontinental ballistic missile force’s launch under attack option based on the results of Montoya’s baccalaureate thesis (2021). Unfortunately, the recent article does not accurately reflect how the United States conducts nuclear deterrence operations.    



We are focusing on three aspects of the larger debate surrounding the advisability of a launch under attack option for the intercontinental ballistic missile force: intercontinental ballistic missile tactics, the accidental launch fallacy, and technical imperatives. We suggest that those, like Montoya and Kemp, arguing for a change in the alert posture of the intercontinental ballistic missile force are mistaken in their assessment of how such a move would affect American deterrence.  

Intercontinental Ballistic Missile Tactics

Montoya and Kemp’s article summarizes the findings of Montoya’s thesis, in which she uses publicly available data to develop simulations of Russian nuclear attacks on America’s intercontinental ballistic missile fields. The pair suggest that at least 100, and possibly up to 200, of the nation’s land-based ballistic missiles would survive a first strike and remain available to the president for a retaliatory strike. 

They argue that the option to launch the American land-based intercontinental ballistic missile force under attack is dangerous, provocative, and unnecessary because these missiles are survivable enough to absorb an attack and have the necessary retaliatory forces required. In fact, they argue that by absorbing a first strike, the advantage shifts to the United States because Russia has used the majority, if not all, of its nuclear forces to destroy only part of the American land-based intercontinental ballistic missile force. 

While not stated directly, the only way to demonstrate a commitment to end the launch under attack option and to prevent the president from executing this option is to de-alert the force. These actions would be dangerous and would undermine America’s response to the rapid nuclear breakout of China and Russian aggression.

Montoya and Kemp are correct in suggesting that it is difficult to successfully eliminate land-based intercontinental ballistic missiles in a first strike because of the total number of weapons required for this task. The 400 missiles, across 450 silos, with 45 launch control centers, and the ability to launch from the Airborne Launch Control System, make the intercontinental ballistic missile leg of the nuclear triad a formidable challenge to a Russian first strike. These characteristics of the nation’s silo-based single-warhead missiles make them valuable; ensuring their destruction is a daunting task that enhances American deterrence.

We  disagree with Montoya and Kemp when they say that “the United States maintains a posture it calls “‘launch under attack,’ a doctrine that permits U.S. missiles to be loosed from their shelters after ‘multiple, independent sensors’ detect an incoming attack from an adversary.” According to the State Department fact sheet they cite, “The United States does not have a launch on warning doctrine [a term used interchangeably with launch under attack].”  

It is also important to distinguish between posture and doctrine. Using the terms correctly is important because force posture has a very specific meaning and is detailed in classified documents such as the “President’s Guidance for the Employment of Nuclear Weapons” and other documents that are used by the Joint Staff, combatant commands, and the services to influence acquisition and operations. Joint and service doctrine establishes how, in the case of the former, the joint force will operate and, in the latter, how each service thinks about and conducts operations. 

Montoya and Kemp are correct in saying that “the United States currently maintains the option to launch under attack so that in the event of a first strike by Russia, U.S. silo-based missiles could be launched before they are destroyed.” An option does not constitute a posture or a doctrine. 

The primary purpose of a launch under attack option is to enhance not missile survivability but deterrence. Deterrence is a psychological effect achieved in the mind of an adversary. The United States enhances deterrence by threatening cost imposition, reducing the benefits of action, and encouraging restraint. Launch under attack reduces the benefits of action by increasing uncertainty and perceived risk. President Putin does not know if he will strike empty silos.  

If the United States de-alerts its intercontinental ballistic missile force, Russia could take a very different tactical approach to striking these forces. After an initial strike, Russia could conduct space-based battle damage assessment using its surveillance and reconnaissance satellites to determine which targets were not destroyed and then restrike surviving launch facilities. As part of a shoot-look-shoot tactic, which is only possible if the missile force is de-alerted, Russia could conduct wave attacks, reducing the number of weapons required on each target in a first strike.  

With a ballistic missile force on alert, Russia must employ a shoot-shoot-look tactic because it must achieve complete destruction with a first strike or risk retaliation. This is necessary because the current launch under attack option forces Russian planners to employ a much higher percentage of the force in a first strike, hoping the United States does not launch its long-range missiles before Russian reentry vehicles strike their targets. This creates the uncertainty needed to deter a first strike. 

It is also worth reiterating that a Russian first strike is highly unlikely prior to a breakout that gives the Russian military significantly more fielded warheads than the United States. In such a situation, launch under attack becomes even more important because a larger Russian arsenal means the percentage of their force needed to conduct a first strike decreases, and exchange ratios are meaningless. 

Montoya and Kemp’s piece ignore the threat of conventional attack on American land-based ballistic missiles. This is surprising because in a 2022 study, Montoya wrote, “One of the most salient issues that surfaced in our study is the projected medium-term increase in the vulnerability of silo-based intercontinental ballistic missile to attack by precision conventional weapons.” Russian performance in Ukraine has underwhelmed observers, but Russia still possesses an arsenal of long-range cruise missiles that can strike anywhere in the United States with accuracy. Since the United States does not field a network of continental air defenses, the homeland is largely unprotected from cruise missile attack. 

Russia is unlikely to use precision conventional weapons to strike missile silos. Instead, cruise or hypersonic missiles are more useful in targeting terrestrial radars, command and control facilities, bombers on the ground, and ported ballistic missile submarines. Such an attack is a way to prevent the United States from building a clear picture of the attack, which makes a response more challenging. The loss of some or most of the bomber and ballistic missile submarine fleet to conventional strike dramatically reduces the president’s response options. If the president also lacks accurate situational awareness from strikes on command and control facilities, the problem is compounded.

With conventional strikes on these other elements of the nuclear triad, Russian intercontinental ballistic missiles can focus on destroying American launch facilities. 

Putin’s recent suspension of Russian participation in New START only underscores our view that any Russian strike on the United States will take place after a breakout that is unmatched by the United States. Given Russia’s track record for cheating on treaties (Convention on Biological and Toxic Weapons, Chemical Weapons Convention, Intermediate-Range Nuclear Forces Treaty, and Open Skies), it is unwise to think they would abide by treaty requirements prior to a nuclear strike on the United States.

The Fallacy of Accidental Launch

In addition to suggesting that a significant portion of the intercontinental ballistic missile force would survive a Russian nuclear strike, Montoya and Kemp argue that “there are many historical examples of early-warning systems generating false alarms or computer-generated messages pretending to be actual warnings. When combined with a launch on warning posture, these glitches create real risks of accidental war.”  

What they fail to mention is that in every example, all of which are decades old, redundant safety measures ensured that any one failure in the system did not lead to an actual failure with nuclear weapons. The U.S. military expects the humans operating the nuclear arsenal and its command and control system to make mistakes. While the military strives for perfection, everything from the weapons themselves to the crews that maintain and operate them are designed to mitigate error.

This is done through the safety measures built into the weapons, training of crews, the personnel reliability assurance program, operational procedures, and a command and control system comprised of layers specifically designed to prevent the very accidental war Montoya and Kemp fret about. In every instance of a mistake or error that detractors can provide, the simple fact is that the redundancy built into the system worked.

In 32 accidents involving nuclear weapons, the United States never experienced an accidental detonation or miscalculation leading to war. Arguments suggesting that because part of the system failed, the entire system failed willfully ignore that the system, which is much better today than when the last accident occurred four decades ago, was specifically designed to account for the inevitable mistakes that would happen.  

The same is true of errors in the systems that comprise American integrated tactical warning and attack assessment. Where one layer failed, another layer succeeded. This layering of systems is sometimes referred to as Reason’s Accident Causation Model, or the Swiss cheese model. There may be holes in one slice of cheese (system), but no hole runs all the way through the entire block of cheese (system of systems). If slices of Swiss cheese are like the layers of redundancy, each slice may have holes in different places, but none of the holes line up perfectly on every slice. Thus, a hole (mistake/error) in one slice is covered in another slice. 

In the aviation world, the crew resource management model builds redundancies into the system to prevent human error when it comes to the combat crews flying nuclear-armed bombers. Similar approaches are in effect across the nuclear enterprise to prevent the kind of accidents Montoya and Kemp fear. Nowhere in the system does safety rely on a single point of failure. Multiple failures must occur, both mechanical or technical and human, before an accidental detonation or nuclear war can happen. It is certainly worth pointing out that no system is perfect. There is always some level of risk, even if it is very small. 

In reality, every human or technical error that occurred in the past was carefully analyzed and used to make the system safer. It is for good reason that the United States has been accident-free for four decades. To continue this safety record, America must invest in people, weapons systems, and nuclear warhead production infrastructure. Regularly building new nuclear warheads that continue to enhance safety and use control is the most reliable way to ensure the least possible risk.         

Technical Imperatives?

Proponents of eliminating the launch under attack option are incorrect to suggest that technological developments are not putting the nation’s ballistic missile submarine fleet at risk. Ballistic missile submarines have a long history of vulnerability to attack when in and leaving port. In 1974, the USS James Madison collided with a Soviet attack submarine sent to stalk U.S. submarines leaving the naval base at Holy Lock in Scotland. Similar incidents occurred when U.S. submarines stalked Soviet submarines leaving their submarine pens. 

High-performance computing is also making it easier to analyze space-based surveillance data with a level of fidelity not possible in past decades. Unmanned underwater vehicles, passive sonar, and other advanced capabilities are also making it harder to hide submarines in the ocean. While submarines are survivable once in their deep ocean boxes, we must continue to invest in keeping them survivable for tomorrow and hedge against technological breakthroughs in antisubmarine warfare.

It is important to keep in mind that the small portion of ballistic missile submarines at sea at any given time are susceptible to attack by conventional torpedo. Submarines in port could face Russian low-observable cruise missile attack. Again, no nuclear strike is needed to decimate the leg of the triad responsible for more than half of all nuclear warheads.  

Since 1991, the bomber fleet has not maintained day-to-day nuclear alert. Both intercontinental and submarine-launched ballistic missiles are armed and ready to launch at a moment’s notice. The bomber fleet must essentially shift from conventional to nuclear operations and move weapons from storage areas to aircraft. This is no easy task. Absent significant warning, the bomber fleet is at considerable risk from both a conventional and a nuclear strike. 

Thus, arguments that suggest an attack on the missile fields are somehow acceptable because the submarine and bomber legs of the triad will go untouched in a conflict are fundamentally flawed. We assess that any attack will begin with attempts to blind the United States by taking out space-based integrated tactical warning and attack assessment capabilities, all while cyber attacks and sabotage attempt to take out command and control. In our assessment, attacks on submarine and bomber bases are also likely to precede or coincide with attacks across the missile fields.  

Military planners must consider the enemy’s most dangerous course of action, in which a Russian attack employs surprise and, consistent with Russia’s operational approach, uses overwhelming force in an initial attack. This leaves the United States insufficient time to deploy the submarine fleet or load and disperse bombers. Under these conditions, ported submarines and much of the bomber fleet are early casualties in a Russian first strike. With the development of a second nuclear-armed peer adversary, America must take the steps necessary to enhance survivability across the triad. 


We do agree with Montoya and Kemp when they write, “Instead of holding fast to the idea of immediate launch, it is far sounder to build a nuclear capability that can survive a first strike and for which decision-makers are not pressed to make decisions with incomplete information.” To achieve this objective, it will take strategic decisions like building mobile intercontinental ballistic missiles, increasing the number of hardened and deeply buried facilities, and placing strategic bombers on dispersed nuclear alert. Continuing on America’s current modernization trajectory will never achieve what both Montoya and Kemp and these authors desire.

It is important to maintain an on-alert missile force capable of launching under attack if the United States desires to deter Russia from contemplating a first strike on the nation’s missile fields. Removing the launch under attack option will not improve the credibility of American deterrence or reduce the risk of accidental detonation or war. It will only further undermine American credibility. With President Putin suspending Russian participation in the New START Treaty, a breakout from treaty restrictions cannot be ruled out. Such a decision would only make a launch under attack option even more important for maintaining deterrence.



Dr. Adam Lowther is Vice President of Research at the National Institute for Deterrence Studies. He spent more than two decades in uniform and as an Air Force and Army civil servant working on nuclear issues. He is also the host of the NucleCast podcast.

Lt. Col. Derek Williams is a B-52 Weapons System Officer and graduate of Sandia National Laboratories’ Weapons Intern program.

The views expressed are those of the authors and do not reflect the official guidance or position of the United States Government, the Department of Defense, the United States Air Force, or the United States Space Force.

Image: U.S. Air Force photo by Senior Airman Clayton Wear