At the height of the Cold War, blurry satellite photos of an obscure nuclear complex in the deserts of Kazakhstan served as a kind of giant Rorschach blot onto which American intelligence analysts projected their worst nightmares.

Some passionately believed that the facility was the center of Soviet efforts to build a particle-beam weapon that could zap American missiles out of the sky. Some thought that the complex was the site of experiments designed to evade a nuclear test ban treaty. And others insisted that the Soviets were developing a new missile.

No one guessed what the rival superpower was really up to. It took the collapse of communism for Soviet scientists to reveal the secret.

While the United States and the Soviet Union frequently misread each other's intentions, there are few more striking examples of the twisted consequences of faulty intelligence than the controversy surrounding the Kazakhstan facility, which was given the acronym P-NUTS by American analysts, for Possible Nuclear Underground Test Site. Paranoia about P-NUTS helped stimulate American research in both particle beam and space-based laser weapons, culminating in President Reagan's decision to launch the multibillion-dollar "Star Wars" program in 1983.

Despite the dire warnings of a Soviet breakthrough in exotic space weapons and the subsequent investment of billions of dollars for research, directed-energy weapons remained only a glint in the eyes of Cold Warriors. Two decades later, the United States has largely abandoned its efforts to develop a functioning beam weapon.

"This is probably the most significant instance during the Cold War of a policy that derived from an incorrect intelligence estimate," says John Pike, a defense analyst at the Federation of American Scientists. "This is a textbook case of satellite imagery being misinterpreted, leading to a huge increase in funding for a specific program."

U.S. expenditures on directed-energy weapons more than doubled under the Carter administration, from $74 million to $161 million, according to Pentagon figures. By the time Ronald Reagan left office in 1989, the country was spending $600 million a year on directed-energy research.

The top-secret satellite photographs that provided much of the initial raw material for the intelligence debate over P-NUTS were recently declassified and made available for research at the National Archives facility in College Park. The pictures show a heavily guarded scientific complex on the southern edge of a nuclear test site near the Kazakh city of Semipalatinsk. Roads crisscross the facility, next to what appear to be long metal pipes and four mysterious-looking holes in the ground.

Other recently declassified documents, and interviews with former intelligence analysts, show that the intelligence community was divided about the purpose of the P-NUTS site for much of the 1960s and '70s. By the end of the '70s, however, there was a growing belief that the site was the focus of Kremlin attempts to gain strategic superiority over the United States by developing a particle beam weapon that would make the Soviet Union invulnerable to nuclear missile attack.

The simplest way of thinking of a particle beam weapon is to imagine a gigantic ray gun--not unlike the ones used by the sci-fi hero Buck Rogers--powered by a nuclear explosion. The difficulties of actually producing such a weapon were immense. First, scientists had to build an underground steel sphere to contain the energy created by the nuclear explosion. Next they had to convert all that power into super-high-voltage electrical energy. Then they had to find a way of storing those billions of volts in a huge battery for a few fractions of a second. Finally, the energy had to be converted into simulated lightning bolts capable of zapping enemy missiles out of the sky.

U.S. scientists had tried to develop particle beam weapons in the early '70s, under a project code-named Seesaw, but ran into overwhelming technical problems at virtually all stages of the research effort. The political climate changed in 1977 when Maj. Gen. George Keegan, a former head of Air Force intelligence, went public with his concerns about a particle beam gap with the Soviet Union.

"This was clearly the genesis of Star Wars," said Pike, referring to the Strategic Defense Initiative championed by Ronald Reagan, which has cost the United States a total of around $50 billion over the past 15 years.

"Keegan's assertions were controversial and far from universally accepted. Nonetheless, they were a significant force in generating the political environment that led the Carter administration to say we needed a larger directed-energy weapons program."

Since the obstacles to building a ground-based particle beam weapon appeared so insuperable, the United States quickly switched the focus of its SDI efforts to space-based lasers. According to Pike, "space-based lasers were our response to our concerns about Soviet particle beam weapons." While particle beam weapons are designed to propel streams of electrons through the air, laser weapons utilize beams of highly focused light. Neither program has yet to produce a functioning weapon.

H. Baker Spring, a research fellow at the conservative Heritage Foundation, agreed that concerns about Soviet technological breakthroughs helped stimulate research here. "Even if we did not understand completely what the Soviets were using this technology for, there was concern that the Soviets were in advance of the United States," he said.

The P-NUTS facility began shedding its aura of mystery in 1992, following the disintegration of the Soviet Union, when Russian scientists invited their American counterparts to tour the complex in the hope of securing Western funding for what turned out to be the real purpose of the facility: a nuclear-powered rocket project. But details of the American intelligence debate over P-NUTS have only recently become available, as a result of declassified government documents and interviews with former U.S. intelligence analysts.

Analysts at the national nuclear laboratory at Los Alamos, N.M., originally concluded that the Soviets were conducting low-yield nuclear weapons tests at P-NUTS. The United States had carried out such experiments in the '50s in an attempt to get around a moratorium on nuclear tests, and it seemed likely that the Soviets were doing the same. The layout of the P-NUTS complex--a series of pipes connected to four large holes in the ground, surrounded by high-security fences--seemed similar to an American low-yield nuclear test facility at Los Alamos.

Fears about particle beam research at P-NUTS were first aroused in the late '60s when satellite pictures showed Soviet workers assembling several steel spheres nearly 60 feet in diameter. The balls--four in all--were then lowered into underground chambers that had been excavated out of rock.

Declassified intelligence documents show that a heated debate broke out within the intelligence community over the purpose of the steel spheres. Convinced that the Soviet Union was way ahead of the United States in Star Wars research, Keegan and his supporters argued that the spheres were designed to contain nuclear explosions that would create the energy required for producing a particle beam "lightning bolt."

Keegan insisted on inserting a dissenting note to the 1976 National Intelligence Estimate produced by the CIA in which he described the Soviet development of particle beam weapons as "the most important strategic undertaking since the development of the atomic bomb." He said the Soviets were "10 to 20 years" ahead of the Americans.

Keegan's rivals in the CIA, who were responsible for the main conclusions of the report, conceded that the Soviets might have conducted some preliminary studies on the feasibility of particle beam weapons. They argued, however, that there was "no convincing evidence" that the weapons were actually under development, and were skeptical of his claim that the P-NUTS site was being used for particle beam research.

Keegan went public with his concerns following his retirement in January 1977, warning in a speech that the Russians might deploy a particle beam weapon as early as 1983. (He did not explain how they would achieve this feat.) Shortly afterward, an article appeared in the influential magazine Aviation Week and Space Technology on the alleged Soviet breakthrough in particle beam weapons.

The article, by Aviation Week military editor Clarence A. Robinson, revealed that U.S. spy satellites had detected the release of large amounts of hydrogen gas from the P-NUTS facility, a telltale sign of the "dumping of energy" that would accompany particle beam weapons tests. The report caused a furor in U.S. national security circles, provoking disclaimers from President Carter, the Pentagon and the CIA. Nevertheless, the controversy provided the political impetus for a significant expansion of the U.S. directed-energy program.

A Harvard graduate who flew 56 combat missions over the Pacific in World War II, Keegan served as chief of Air Force intelligence for five years at the height of the Cold War. Convinced that the Soviet Union was secretly preparing for war against the United States, he consistently challenged official estimates of Soviet military power, taking an even darker view of Soviet intentions than most of his colleagues in the intelligence community. (He accused the CIA of providing employment to "left-leaning" analysts intent on nuclear disarmament.)

Keegan's warnings of Soviet strategic superiority quickly turned him into a hero of the far right. But his estimation of Soviet power--that the Soviets could overrun Western Europe in a day and a half and sink 75 percent of the U.S. Navy in a matter of hours--were derided by other experts.

Of all Keegan's obsessions, fear that the Kremlin was about to develop a particle beam weapon was perhaps the most consuming. He taught himself nuclear physics while still chief of Air Force intelligence in order to validate his theories. In 1978 he told CBS's "60 Minutes" that the Soviet Union had embarked on "the most gigantic scientific program of its kind in history" and that "time was running out for the United States." Standing in front of an artist's rendition of the particle beam project allegedly underway in Kazakhstan, Keegan described how the Kremlin was working on weapons that would "simply eviscerate" incoming U.S. warheads.

"We are talking about billions of electron volts of energy," said Keegan, explaining how the nuclear energy produced inside the steel balls would be converted into "streams of electrons" through the use of an "electron gun injector." Keegan died in March 1993, just as his theories about Soviet military superiority were becoming widely discredited. There is no evidence that he rethought his views.

By going public with his warnings, Keegan may have exerted a much greater influence on the intelligence debate than he ever had in private. Recently declassified National Intelligence Estimates show that his views about particle beam weapons eventually gained acceptance. By the end of 1977, for example, the CIA had reversed its earlier skepticism and reported that "the evidence"--precisely what evidence is unclear--now suggested that the Soviets were indeed engaged in extensive particle beam weapons research.

Russian scientists who escorted American visitors through the P-NUTS site in 1992 insisted that their goal was merely a nuclear-powered rocket to Mars. They said that the gigantic metal spheres that attracted Keegan's attention were used to store the hydrogen that would propel such a rocket after being heated in a nuclear reactor, a claim consistent with the U.S. intelligence finding of the release of large amounts of hydrogen from the site. Most American scientists now accept the Russian description of the purpose of the Semipalatinsk test site.

"Everybody was amused by the [particle beam] rumors," said Elliot Kennel, a research physicist from Cedarville, Ohio, who toured the P-NUTS plant in 1992. He remembers walking along one of the underground tunnels that Keegan believed was a channel for a particle beam, and joking with his Soviet guide about having to duck if the weapon was suddenly switched on. After personally inspecting the underground steel spheres, Kennel concluded that they were not nearly strong enough to contain a nuclear explosion, which would have been necessary to generate the energy for a particle beam weapon.

"What we saw was a very unique, probably best-in-the-world, facility for nuclear space propulsion," said Leonard Cavany, a retired government physicist who worked on the Strategic Defense Initiative and who went on the 1992 tour.

Still, Robinson, who broke the particle beam story for Aviation Week in 1977, says that he has "no reason" to doubt the accuracy of his reporting. "I don't know what has happened in the last decade, but I am satisfied that what I wrote was believed to be accurate by the intelligence community at that time." He quotes a former CIA chief as telling him in the early '90s that Keegan was "more right than wrong."

But there is little doubt that official America accepted the on-the-ground reality that P-NUTS was ultimately about making better spaceships, not better weapons. After the end of the Cold War, the Bush administration even gave serious consideration to a joint project with Russia on nuclear propulsion for a possible Mars cargo mission based on research already done at the Semipalatinsk facility, though budget constraints caused the Clinton administration to mothball the project.

So it has all come to nothing, an odd footnote in the history of a bygone era. But it is one of the more interesting ironies of Cold War paranoia that the debunking of America's belief in a Soviet death ray came as a big disappointment--to Russian scientists.

Kennel was told by the Russians on his tour that in the last years before the fall of communism researchers arrived at the Semipalatinsk plant primed by foreign media reports and rumor. They could hardly wait to begin working on the death ray, popularly referred to among themselves as "Keegan's beam."

When they discovered the truth, morale suffered. The idea of building nuclear-powered rockets must have seemed pretty tame. After all, they had been expecting to hurl lightning bolts.