Weapons That Did Not Make the Cut

The U.S. Nuclear Weapons Cost Study Project was completed in August 1998 and resulted in the book Atomic Audit: The Costs and Consequences of U.S. Nuclear Weapons Since 1940 edited by Stephen I. Schwartz. These project pages should be considered historical.

Twenty-five weapons programs utilizing sixteen warhead types were canceled before production, because another warhead type was chosen, or in some cases, because the delivery system itself was canceled (see table 2-4). Each canceled program makes an interesting story, capturing as it does the thinking of the day. Some warhead types have had wide applicability, used in one configuration as a bomb, and in another as a warhead for one or perhaps several kinds of missiles. These delivery systems include the Navaho cruise missile ($4.9 billion); the Skybolt, an air-launched ballistic missile ($2.6 billion); the Midgetman or Small ICBM ($5.6 billion); the MX Rail Garrison basing plan ($3.4 billion); the SRAM II ($1.1 billion); and the PLUTO, a nuclear-powered, nuclear-armed, low-altitude cruise missile (about $780 million).


From the vantage point of the early 1950s, it appeared that the cruise missile, rather than the ballistic missile, would be the first to be deployed.


The Navaho was a supersonic cruise missile whose roots could be traced to the early postwar American efforts to modify and upgrade the German V-1. By September 1950 the Air Force had proposed a three phase program consisting of the design and construction of a test vehicle (plus a series of test flights), the development of an interim missile with a range of 3,600 miles (5,792 kilometers), and, finally, the development of an operational missile with a range of 5,500 miles (8,850 kilometers).

The test vehicle, called the X-10, used two Westinghouse turbojets to power the 70-foot-long (21 meter) missile. Eleven vehicles flew 27 flights beginning in October 1953. In one test, the X-10 reached a speed of Mach 2.05. But the interim missile, known as the XSM-64, had repeated launch failures during 1956-1957. With competition from the more promising ballistic missile programs under concurrent development, the Air Force canceled the Navaho on July 12, 1957. A limited number of test flights after cancellation, five of which occurred between August 12, 1957 and February 25, 1958, with two final tests taking place on September 11, 1958 and November 18, 1958. For a total of one and a half hours of flight time, U.S. taxpayers spent around $4.9 billion. The Navaho itself never flew, although some of the technology developed for it was used in subsequent missile and space programs. The proposed intercontinental XSM-64A was 87.3 feet (26.6 meters) long, with a wing span of 40.2 feet (12.3 meters) and weighed 120,000 pounds (54,545 kilograms) at launch. There was also a 91.5-foot-long (27.9-meter), 169,500-pound (77,045-kilogram) booster to which the Navaho would ride piggy back for 110 seconds.

The Navaho, like the


(another cruise missile, thirty of which were deployed for four months at Presque Isle, Maine, at a cost of $4.2 billion), failed for a number of reasons.


The technology of the day could never match the ambitious requirements set by the Air Force. Loose corporate management led to cost overruns and schedule slippage. The ballistic missile, though unproven as well, was winning the competition (it was more accurate and reliable, and had a shorter flight time) and money, resources, and attention turned to it.

Table 2-4. Canceled U.S. Nuclear Capable Missile Programs, 1943-91

        Total program cost
Missile Service Warhead/canceled Start-end Then-year Constant 1996a
Rigel SSM (SSM-N-6) Navy W5/August 1953 1943-53 38 380
Hermes SSM (A-1) Army W5/August 1953 1944-54 96 970
Sparrow I AAM (AIM-7A) Navy W42/January 1958 1945-56 196 1,370
Rascal ASM (AGAM-63) Air Force W5/March 1956 1946-58 448 3,240
Triton (SSM-N-2) Navy W27/December 1955 1948-57 19 140
Navaho SSM (SM-64) Air Force W13/September 1954 1954-57 680 4,910
Corvus ASM-N-8 Navy W40/August 1960 1954-60 80 550
Regulus II RGM-15A Navy W27/? 1952-58 147 1,120
Nike Zeus Army W50/? 1955-65 3,000 18,360
Honest John, Jr. SSM Army W52/February 1956 ? ?  
Crossbow ASM (GAM-67) Air Force W31/January 1957 1957-58 75 520
Hopi ASM Navy W50/December 1958 ? ?  
Hawk SAM Army W42/December 1960 ? ?  
Typhon SAM (RIM-55) Navy W60/March 1964 1958-64 225 1,370
Eagle AAM-N-10 Navy W42/c. March 1961 1959-61 53 320
Pluto Air Force ? 1959-64 128 780
Mobile Minuteman Air Force ? 1959-62 108 660
Skybolt AGM-48A Air Force W59/December 1962 1960-63 440 2,640
MMRBM Air Force ? 1962-64 65 380
Condor ASM Air Force W73/September 1970 ? ?  
Hound Dog II Air Force ? 1972-73 12 40
SRAM II AGM-B1A Air Force W89/1990 1983-90 858 1,020
SRAM-T AGM-B1B Air Force W91/September 1991 1983-90 81 110
SICBM MGM-134A Air Force W87/1991 1984-91 3,675 5,600
MX/Rail Garrison Air Force W87/1991 1986-91 2,088 2,400
Total       12,512 46,880


: Thomas B. Cochran and others,

Nuclear Weapons Databook,

vol. 2:

U.S. Nuclear Warhead Production

(Cambridge, Mass.: Ballinger, 1987); Nicholas and Rossi,

U.S. Missile Data Book,

pp. 3-11-3-13; U.S. Army Missiles Handbook (Office of Director of Progress and Statistical Reporting, Office Comptroller of the Army, OCS, January 15, 1960); Donnelly, "The United States Guided Missile Program."


Except for Skybolt, MMRBM, SRAM II, SICBM, and MX Rail Garrison, for which verified annual data exist, constant 1996 dollar totals were computed by using the midpoint of development to select an appropriate deflator (for example, a system in development from 1950 to 1960 would be adjusted using a 1955 deflator). The resulting figure was then rounded to the nearest $10 million. This is obviously a very rough calculation, but barring the discovery of annualized budgetary data for most of these programs, it is the most reasonable means available.



Skybolt was an effort by the Air Force to utilize B-52 bombers as ballistic missile launchers.


As originally conceived, Skybolt was to have been a 39-foot (11.9-meter), 11,000-pound (5,000-kilogram), two-stage missile with a range of some 950 nautical miles (1,759 kilometers) when dropped from a B-52 at a height of 40,000 feet (12,192 meters). At maximum range, it would reach its target in approximately twelve minutes. Other launch heights and ranges were also considered. Its mission, developed in an effort to regain ground lost with the introduction of the Navy's new Polaris missile and submarine, was "defense suppression": destroying air defense batteries to allow SAC bombers clear paths to Soviet targets (that the Polaris, then being deployed, could already accomplish this task did not deter the Air Force from proceeding). The Air Force originally sought to purchase 1,000 missiles to equip twenty-two bomber squadrons by mid-1967 for $15.3 billion, of which $3.7 billion was for the missile's planned 800-kiloton warhead.

At the urging of the Air Force, the DOD on February 1, 1960, authorized the Skybolt R&D program. That March, the British government, eager to bolster its independent nuclear force, expressed an interest in using the Skybolt in conjunction with its Vulcan bombers. It received assurances from President Eisenhower that such a program would be developed and formed a joint project office to work with the Air Force (in a secret quid pro quo deal agreed to during a meeting that month at Camp David, British Prime Minister Harold Macmillan offered Eisenhower the use of Holy Loch, Scotland, as a strategic submarine base).

The Skybolt testing program, which included building and dropping full-size dummy missiles from both U.S. and British aircraft, failed to achieve much success and, with costs mounting, President Kennedy canceled it in December 1962, disappointing the Air Force, which had, ironically, been offered the project almost two years earlier as "compensation" for Kennedy's cancellation of the troubled B-70 bomber,


and straining relations with the British. Kennedy was prepared to offer Britain the data and materials from the Skybolt program in exchange for a $100 million (then-year dollars; $588 million in 1996 dollars) cash payment to cover development costs. But with Skybolt deemed a failure, British Prime Minister Harold Macmillan and his Defense Minister, Peter Thorneycroft, set their sights higher and eventually persuaded Kennedy (against the advice of the State Department, which was deeply concerned about the example this would set) to offer cooperation on the Polaris SLBM program.


In all, Skybolt cost nearly $2.6 billion by the time it was terminated.

Project PLUTO. Project PLUTO

sought to combine the power of nuclear weapons with the presumptive versatility of a flying nuclear reactor by creating the world's first atomic-powered buzz bomb. Using a nuclear-powered ramjet engine and carrying a nuclear warhead, the PLUTO missile would have flown at supersonic speeds around treetop level (to avoid radar detection) and delivered its lethal cargo without need of refueling or a crew to guide it. The Air Force, AEC, and JCAE were all strongly supportive of the concept, which went as far as ground-testing of its turbineless engine at the Nevada Test Site in 1961.

But even as the concept was proving itself feasible, military officials began to have doubts about it. One basic problem was that the ramjet - like the engine then in development for the nuclear-powered aircraft - was essentially an open-ended nuclear reactor, meaning that even routine operations would spew vast quantities of dangerous fission products in its wake, to say nothing of the potential for in-flight accidents or crashes. When combined with a nuclear warhead, the need to transit allied territory on the way to its targets in the Soviet Union, and the missile's vulnerability to being shot down PLUTO became less and less attractive. Moreover, the radioactive fallout generated by the vaporization of PLUTO's reactor engine would have been many times greater than for just the warhead alone, creating long-term risks downwind of any explosion site. The program was canceled in 1964, following expenditures of about $780 million.


Project Orion.

One other project deserves mention for it demonstrates the lengths scientists and engineers went to in attempts to utilize the power of nuclear weapons. Project Orion, was a proposal to power a spacecraft by detonating nuclear bombs behind it, allowing the force of the blast to hit a specially-designed plate on the back of the craft and thus propel it forward. Despite the obvious problem of safely launching such a vehicle from Earth, scientists worked up a number of studies of the concept, one of which modeled the device to eject the bombs from the craft on the bottle dispensing mechanism in Coca-Cola vending machines.


From 1958 until early 1965 the studies consumed nearly $50 million. In 1960, the program was transferred from the Advanced Research Projects Agency (ARPA) to the Air Force which, seeing little weapons potential in it, appeared to abandon it. However, a recently declassified memorandum to General Curtis LeMay, dated June 9, 1964, indicates that the concept was kept alive through at least that date by the Air Force, General Atomics (a defense contractor) and scientists at Livermore Laboratory, who had transformed it into Project Helios, proposing to use lower yield nuclear explosives to propel a smaller, lighter spacecraft from orbit.


Scientists at Livermore suggested proving the new concept via underground nuclear tests, an approach the Air Force Scientific Advisory Board favored so long as NASA and the AEC would take a direct financial interest in the project. The outcome of this proposal is unclear, but Orion never took flight.

A separate but related effort, known as Project ROVER (and later Nuclear Engine for Rocket Vehicle Applications, or NERVA) sought to build nuclear reactor engines to power missiles and rockets for both military and space exploration applications. Initially considered a possible means of powering ICBMs, the program later concentrated on their use as the second-stage motor of a lunar mission and for manned missions to Mars. In all, twenty-one reactor tests took place at the Nuclear Rocket Development Station at the Nevada Test Site between 1959 and 1969. On January 21, 1965, a Kiwi-B type reactor was deliberately destroyed by allowing it to blow apart in order to assess analytical models of the reactor's behavior during a rapid power excursion (only 50 percent of the reactor core could be accounted for after this test; the remainder was presumed to either have burned or been converted into fine particles and carried downwind in a cloud). The program was terminated following a shift in national priorities, principally a move away from large numbers of manned spaceflights following the success of the Apollo program. From 1961 to 1973, NERVA consumed $3.9 billion.





Evolution of the Cruise Missile,

pp. 97-108; U.S. Air Force, Air Research and Development Command,

Development of the SM-64 Navaho Missile, 1954-1958

(Historical Branch: Wright Air Development Division, January 1961). For a technical history of the program, see James N. Gibson,

The Navaho Missile Project: The Story of the ?Know-How' Missile of American Rocketry

(Atglen, Pennsylvania: Schiffer Publishing., 1996).



Of Snark, Dr. Herbert York (the former director of defense research and engineering) would later write, "We realized, several years before Snark became operational, that it would become obsolete by the time it was finally deployed, and repeated recommendations for dropping the project were made. However, in this case as in so many others, the momentum of the project and the politics which surrounded it made it impossible to do so." Herbert F. York,

Race to Oblivion: A Participant's View of the Arms Race

(New York: Simon and Schuster, 1970), p. 80. General Thomas Power, head of the Strategic Air Command, stated in 1959 that, "it is a subsonic missile. Let's face it, it came late. We are buying only a very few of them. They have a very limited value to SAC." Quoted in Beard,

Developing the ICBM,

p. 212, n. 30. As Beard notes, "It is possible, of course, that one factor at play in this period was a hesitancy to cancel Snark and Navaho on grounds of vulnerability because such arguments might also reflect poorly on large segments of the manned fleet." Beard,

Developing the ICBM,

p. 212, n. 30. The first Snark at Presque Isle went on alert on March 18, 1960 and on February 28, 1961, SAC declared the base fully operational with 30 missiles. One month later, President Kennedy ordered the Snark to be phased out because it was "obsolete and of marginal military value." By June 25, 1961, all Snarks were withdrawn from service. Gibson,

History of the US Nuclear Arsenal,

p. 148.



Ian Clark,

Nuclear Diplomacy and the Special Relationship: Britain's Deterrent and America, 1957-1962

(Oxford: Clarendon Press, 1994), pp. 338-373; Richard E. Neustadt,

Alliance Politics

(Columbia University Press, 1970), pp. 30-55; Robert Standish Norris, Andrew S. Burrows, and Richard W. Fieldhouse,

Nuclear Weapons Databook

vol. 5:

British, French, and Chinese Nuclear Weapons

(Boulder, Colo.: Westview Press, 1994), pp. 98-99.



Kennedy and McNamara offered this "compensation" despite strong reservations about the viability of Skybolt. After making the decision to terminate the bomber, Kennedy is reported to have remarked that he, "used Skybolt to shoot down the B-70." See York,

Race to Oblivion,

p. 155.



McNamara was prepared to offer the British the Hound Dog cruise missile, but this provoked disappointment and offense. How, Neustadt notes (citing Henry Brandon), "could Englishmen base ?independence' upon something labeled Hound Dog?" Neustadt,

Alliance Politics,

pp. 47-48.




The Atomic Establishment,

pp. 201-03; Letter from Representative Chet Holifield, Chairman, Joint Committee on Atomic Energy, to McGeorge Bundy, Special Assistant to the President, December 22, 1962, Record Group 128, General Correspondence, Box 668 (Project Pluto), National Archives. In this letter, Holifield reports "expenditures] of $128 million" in then-year dollars for PLUTO. However, available official budget data record only $94.3 million (then-year) in obligations through fiscal 1963 (total recorded obligations through 1965 were $110.3 million). Because relatively small programs like PLUTO were not always itemized in annual budgets and assuming that Holifield, a strong proponent of the program and at the time chairman of the JCAE, had access to all pertinent data, we choose to use his figure here, adjusting it to constant 1996 dollars using a base year of 1961.



John McPhee,

The Curve of Binding Energy,

(New York: Farrar, Straus and Giroux, 1974), p.



Memorandum, James Ferguson, Lieutenant General, U.S. Air Force, Military Director, USAF Scientific Advisory Board, to General Curtis E. LeMay, June 9, 1964, Curtis LeMay papers, Library of Congress.



W.H. Robins, Analytical Engineering Corp., and H.B. Finger, consultant, "An Historical Perspective of the NERVA Nuclear Rocket Engine Technology Program," paper presented at the AIAA/NASA/OAI Conference on Advanced SEI Technologies, September 4-6, 1991, Cleveland, Ohio, AIAA (American Institute of Aeronautics and Astronautics) 91-3451; Daniel R. Koenig, "Experience Gained from the Space Nuclear Rocket Program (Rover)," Los Alamos National Laboratory, LA-10062-H, May 1985; Gary L. Bennett and others, "Prelude to the Future: A Brief History of Nuclear Thermal Propulsion in the United States," Special Commemorative Paper, NASA Symposia on Space Nuclear Power Systems, Preprint Log 092 (n.d., circa 1992). Budget data for ROVER/NERVA were obtained from successive volumes of Bureau of the Budget/Office of Management and Budget,

Budget of the United States Government,

Fiscal Year 1958 through 1975, and converted to constant 1996 dollars.


Copyright © 1998 The Brookings Institution