Is Remote Viewing Real? The CIA’s Stargate Project and What the Evidence Actually Shows

For twenty-three years, the United States government funded a program to see if trained psychics could spy on enemy targets using nothing but their minds. The program produced extensive classified reports, statistical analyses from professional statisticians, and a conclusion that split the scientific community in half. Here is what actually happened, what the evidence shows, and why it still matters.

—

The Cold War Origin Story

In 1970, the CIA received intelligence reports suggesting the Soviet Union was spending millions, eventually estimated at 60 million rubles per year, on “psychotronic” research. The Soviets were reportedly investigating telepathy, psychokinesis, and clairvoyance as potential intelligence tools. Whether this intelligence was accurate remains debatable, but the CIA’s reaction was not: they decided they needed to know if psychic phenomena were real, and if so, whether the Soviets might weaponize them before the United States could understand them.

This was the Cold War. The United States had already funded research into LSD as a potential mind-control agent (MKUltra), remote-controlled insects (Project Bluebird), and assassination methods involving exploding cigars. Investigating whether humans could perceive distant locations through extrasensory means was, by comparison, a relatively modest expenditure of paranoia.

The CIA’s initial contractor was SCANATE (“scan by coordinate”), launched in 1970. The program went through a string of code names as it migrated between agencies: Gondola Wish (Army Intelligence, 1977), Grill Flame (INSCOM, 1978), Center Lane (1983), Sun Streak (DIA, 1985), and finally Stargate (1991). Each rename reflected a handoff, a budget restructuring, or a new office trying to distance itself from whatever went wrong under the last name. By the time it was shut down in 1995, the total U.S. investment across all program iterations was approximately $20 million.

The CIA contracted two physicists at Stanford Research Institute (SRI) in Menlo Park, California: Russell Targ, a laser physicist who had worked on early laser radar systems, and Harold “Hal” Puthoff, a quantum physicist who had published extensively on quantum electronics. Both were credentialed scientists with mainstream publication records. Neither was a psychic, a mystic, or a paranormal enthusiast. They were engineers asked to test a hypothesis.

Key Researchers

Researcher	Role	Affiliation	Stance
Russell Targ	Co-founder, SRI remote viewing program	SRI International	Proponent
Hal Puthoff	Co-founder, SRI remote viewing program	SRI International	Proponent
Edwin C. May	Program director (1976-1995)	SRI, later SAIC	Proponent
Joseph McMoneagle	Primary remote viewer	U.S. Army Intelligence	Proponent (participant)
Jessica Utts	Statistical evaluator (AIR, 1995)	UC Davis	Proponent
Ray Hyman	Methodological evaluator (AIR, 1995)	University of Oregon	Skeptic
James Randi	Independent investigator	Magician/investigator	Skeptic

—

The Jupiter Session

Ingo Swann is a figure who defies easy categorization. An artist by training, he is credited with coining the term “remote viewing” in December 1971 during an experiment at the American Society for Psychical Research in New York. He also designed the Coordinate Remote Viewing (CRV) protocol, a structured six-stage method that replaced the earlier “close your eyes and tell me what you see” approach with something closer to a laboratory procedure.

The CRV protocol was not a personality test. It was a system designed to make the research reproducible. The six stages moved from raw sensory impression to structured analysis:

Stage I: Ideograms and gestalt impressions (the viewer’s hand produces a reflexive drawing when given coordinates)
Stage II: Sensory descriptors (textures, temperatures, colors, sounds at the target site)
Stage III: Dimensional sketches (rough drawings of the site’s layout)
Stage IV: Detailed analysis (the viewer processes accumulated data into coherent descriptions)
Stage V: Queries and hypotheses (the viewer identifies specific questions about the target)
Stage VI: Modeling and 3D representation (the viewer builds a physical model of the target using clay or drawings)

The system was later refined by Tom McNear, the only person trained in all six CRV stages by Swann. Whether CRV actually works is a separate question from whether it represented a genuine attempt at making parapsychology more rigorous. It did represent that attempt.

But the session that made Swann famous, the one that still generates arguments, took place on the evening of April 27, 1973, at SRI. Targ and Puthoff asked Swann to remote view the planet Jupiter.

This was not a controlled experiment in the formal sense. Jupiter is not a location you can verify by driving there. But the session produced a detail that would not be confirmed until 1979, when the Voyager 1 spacecraft flew past Jupiter: Swann described rings around the planet.

Jupiter’s rings were not known to exist in 1973. The first confirmed observation came from Voyager 1’s photopolarimeter on March 4, 1979, six years after Swann’s session. The rings are faint, composed of dust, and invisible from Earth with the telescopes available in the 1970s.

This is the kind of anecdote that either convinces you or doesn’t, depending on what you already believe. Proponents point to it as a dramatic, unambiguous hit. Skeptics note that Swann produced pages of material about Jupiter, most of it wrong or unverifiable, and that describing “rings” around a planet is a low-cost prediction when Saturn’s rings were already well known. If someone asked you to describe an unfamiliar room, you might say “there’s probably a table”. Not because you’re psychic, but because most rooms have tables.

The truth is somewhere in between. The detail was specific, it was recorded before confirmation, and it was correct. It is also a single data point from an uncontrolled session, which is not how you build a scientific case. But it is how you build a legend.

—

The Evidence for Remote Viewing

The Original SRI Experiments (1972-1979)

Targ and Puthoff’s initial experiments followed a simple design: a target location was randomly selected from a set of possibilities. A “viewer”, a person claiming to have psychic ability, would describe the target while the experimenter recorded their verbal descriptions. An analyst would then compare the descriptions to the actual targets and score the matches.

Their earliest test subjects included Uri Geller, the Israeli entertainer who would later become internationally famous for bending spoons on television. Targ and Puthoff reported that Geller could draw pictures sealed in opaque envelopes that he never touched. Ray Hyman, sent by ARPA to evaluate the results, concluded that Geller was a “complete fraud.” Targ and Puthoff lost their government contract for further work with Geller. The initial interest their research had generated within the defense establishment survived the relationship, and the program continued with other subjects.

One of those other subjects was Rosemary Smith, a young administrative assistant recruited by project director Dale Graff. In 1976, Smith reportedly helped locate a lost Soviet spy plane. Not through classified satellite imagery, but through remote viewing. The case was cited by proponents as evidence that the method could produce actionable intelligence. Whether the plane’s location was genuinely unknown when Smith viewed it, or whether some conventional intelligence channel had already narrowed the search area, remains a matter of debate.

Their first published paper, “Information Transfer Under Conditions of Sensory Shielding,” appeared in Nature in 1974. The paper reported that viewers could describe randomly selected target sites with accuracy significantly above chance. 🟡

Funding disclosure: Targ & Puthoff (1974): funded by SRI International internal funds and CIA contract. Researchers were employees of the funding agency. Publication venue: Nature (top-tier peer-reviewed journal).

The Nature publication was significant. Getting a parapsychology paper into one of the world’s most prestigious scientific journals was, by itself, an extraordinary event. Critics would later argue that Nature‘s peer review process was not designed to detect subtle methodological flaws in parapsychological experiments, but the publication gave the research immediate mainstream visibility.

The SAIC Period and the AIR Evaluation (1991-1995)

By 1991, the program had moved from SRI to Science Applications International Corporation (SAIC), with Edwin C. May as director. The SAIC experiments were more rigorous than the original SRI work, using double-blind protocols, automated target selection, and blinded judging.

In 1995, the CIA contracted the American Institutes for Research (AIR) to evaluate the program before deciding whether to continue funding it. AIR hired two independent evaluators with opposing viewpoints: Jessica Utts, a statistician at UC Davis known for being open to parapsychological evidence, and Ray Hyman, a psychologist at the University of Oregon known for skeptical analysis.

Utts’ conclusion: “Using the standards applied to any other area of science, it is concluded that psychic functioning has been well established. The statistical results of the studies examined are far beyond what is expected by chance.” 🟢

Hyman’s conclusion: Regarding Utts’ statistical analysis, Hyman wrote that her conclusion that ESP had been proven to exist “is premature, to say the least.” He argued that the laboratory results, while statistically anomalous, could not be taken at face value due to unresolved methodological concerns, including the lack of a replicable protocol and the possibility of subtle sensory leakage in some experiments. 🟡

Funding disclosure: Utts (1996): funded by CIA/AIR evaluation contract. Utts was a paid evaluator, not a program participant. Hyman (1996): same contract. Both evaluators were independent of the research team.

The critical point that is often missed: both evaluators agreed that a statistically significant effect existed in the data. They disagreed on whether the methodology was rigorous enough to trust the effect as real.

The Cardeña Review (2018)

In 2018, Etzel Cardeña published a comprehensive review in American Psychologist (the flagship journal of the American Psychological Association). Cardeña reviewed decades of experimental evidence across multiple psi phenomena, including remote viewing.

“The evidence provides cumulative support for the reality of psi, which cannot be readily explained away by the quality of the studies, fraud, selective reporting, experimental or analytical incompetence, or other frequent criticisms. The evidence for psi is comparable to that for established phenomena in psychology and other disciplines, although there is no consensual understanding of them.”

Funding disclosure: Cardeña (2018): no listed conflicts. Publication venue: American Psychologist (top-tier peer-reviewed journal). 🟢

This review is notable because American Psychologist is not a fringe journal. It is the APA’s own publication, and its acceptance of this review represents a significant mainstream acknowledgment that the evidence base exists, even if interpretation remains contested.

—

The Case Against

Intellectual honesty requires presenting the strongest objections. Here is what the skeptical literature says about remote viewing, and where the critique is strongest.

1. The Effect Sizes Are Tiny

The most damaging criticism is simple: even in the studies that found statistically significant results, the effect sizes were small. When Ray Hyman reviewed the SAIC data, he noted that “the overwhelming amount of data generated by the viewers is vague, general, and way off target.” The statistically significant hits were often buried in volumes of misses. A viewer might produce 50 pages of description, of which 3-4 sentences happened to match the target. The statistical significance comes from aggregation across many trials, not from dramatic individual sessions.

This is a legitimate concern. An intelligence officer needs actionable, specific information, not a statistical anomaly detectable only across thousands of trials. As the AIR evaluators concluded, remote viewers “consistently failed to produce actionable intelligence information” (AIR evaluation, 1995).

2. Replication Has Been Inconsistent

Hyman’s core criticism of the Utts evaluation was that “the findings had yet to be replicated independently.” While multiple labs have reported remote viewing effects, the replication record is not clean. Some labs have found effects, others have not, and the conditions under which replication succeeds or fails are poorly understood.

The PEAR (Princeton Engineering Anomalies Research) laboratory, which ran experiments from 1979 to 2007, reported statistically significant effects from random number generator experiments. However, the effect sizes were even smaller than the SRI/SAIC studies, and critics argued that the PEAR results, while statistically anomalous, were practically meaningless.

3. No Plausible Mechanism

Physics as currently understood provides no mechanism by which a human brain could perceive a distant location without any sensory input. Remote viewing would require information to travel from a target to a viewer through some channel that does not correspond to any known physical force, not electromagnetic radiation, not sound, not gravitational effects. Proponents have suggested quantum entanglement or nonlocal consciousness, but these are speculative analogies, not established mechanisms.

This is not disproof, the history of science includes many phenomena observed before mechanisms were understood (electricity, continental drift, dark matter). But it does mean the bar for evidence should be higher, not lower, because the claim is extraordinary.

4. The Confirmation Bias Problem

The remote viewing research community is small. The same researchers (Targ, Puthoff, May, and their collaborators) produced the majority of positive results. The citation network is concentrated: most papers cite the same core group of studies. This creates a risk of circular validation, where a small community of researchers cite each other’s work, creating an impression of broad support that is actually narrow.

Additionally, some of the researchers had a vested interest in the program continuing. Their careers, reputations, and funding depended on the program’s continuation. This doesn’t invalidate their results, but it is a conflict of interest that should be noted.

5. The “Successful Failures” Problem

When remote viewing sessions produced accurate descriptions, they were celebrated. When they failed, which was most of the time, the failures were attributed to viewer fatigue, environmental noise, or methodological issues. This asymmetric treatment of results is a hallmark of confirmation bias. If successes count as evidence for psi and failures don’t count as evidence against it, the framework is unfalsifiable.

6. The NRC’s Verdict

In 1988, the National Research Council published “Enhancing Human Performance,” a comprehensive review of techniques proposed to improve human capabilities, everything from sleep learning to parapsychology. Chapter 9 covered paranormal phenomena.

The NRC’s conclusion about the Targ-Puthoff SRI experiments was blunt: “There should remain little doubt that the Targ-Puthoff studies are fatally flawed.” The panel found that the studies lacked adequate controls, that sensory leakage could not be ruled out, and that the reported results were not replicable under conditions that eliminated these confounds.

This is the kind of verdict that kills programs in normal science. A government-commissioned review by the National Academy of Sciences says your work is fatally flawed, and you lose your funding. But the Stargate program survived the NRC report, partly because the military intelligence community had already invested too much to walk away, and partly because the NRC’s review covered only the early SRI experiments. The later work at SAIC was more rigorous and produced its own, independent statistical results.

7. The Uri Geller Problem

Early in the SRI program, Targ and Puthoff tested Uri Geller and reported remarkable results. Ray Hyman, sent by ARPA to evaluate, called Geller a “complete fraud.” The researchers lost their government contract for further work with Geller.

The Geller episode is more than an embarrassment. It illustrates a pattern that plagued the entire program: the researchers’ enthusiasm for positive results sometimes outpaced their skepticism about the subjects producing them. If your first major test subject is later discredited, it raises questions about the selection criteria for every subject who followed. McMoneagle may have been the genuine article, but how would you know the difference?

—

Implications

Level 1: What It Means for Science

Whether or not remote viewing is real, the Stargate program demonstrated that the scientific establishment has a complicated relationship with anomalous claims. The evidence base for psi is not zero. It exists in peer-reviewed journals, was evaluated by professional statisticians, and was funded by government agencies with access to classified information. Dismissing it entirely requires dismissing the peer review process that accepted it.

But the evidence is also not conclusive. The effect sizes are small, replication is inconsistent, and no mechanism has been identified. The honest position is: the evidence is suggestive but insufficient.

The broader lesson is about how science handles anomalies. The psi literature shares features with other contested research areas: small effect sizes, replication failures, and methodological disputes that have emerged in mainstream psychology during the replicability crisis. Some of the same statistical techniques (p-value thresholds, small sample sizes, researcher degrees of freedom) that made psi studies controversial have since been found to produce false positives in social psychology. The psi debate may have been an early warning of problems the field as a whole would later confront.

Level 2: What It Means for Government Research

The Stargate program was not shut down because it was debunked. It was shut down because the effect was too small to be operationally useful. The AIR evaluators concluded that remote viewing information was “too unclear and non-detailed” to be used in any intelligence operation. This is an important distinction: a statistically significant laboratory effect does not automatically translate to a useful tool.

The program’s 23-year run also demonstrates something about institutional inertia in intelligence funding. A program that showed marginal results in controlled conditions continued to receive government money for over two decades, partly because the Soviet threat made the potential upside seem enormous, and partly because intelligence agencies have a structural inability to say “we spent money and got nothing.” The sunk cost fallacy is not just a cognitive bias. In government, it’s a career strategy.

Level 3: What It Means for Individual Practice

The question people actually want answered: “Can I learn to remote view?” is a question not addressed by the laboratory studies. The research tested whether the phenomenon exists under controlled conditions, not whether individuals can develop it as a skill. The Stargate program used a small number of trained viewers (primarily Joseph McMoneagle), and the transferability of the ability to the general population is unknown.

Former president Jimmy Carter provided one of the most frequently cited anecdotes. In a 2005 interview with GQ magazine, Carter recalled that during his presidency, a woman in California had reportedly used psychic ability to locate a small twin-engine plane that had gone down somewhere in Africa. Carter said the CIA director confirmed the plane’s location through this method. Carter himself responded “with skepticism,” saying “Whether it was just a gross coincidence or… I don’t know.” The anecdote is remarkable because it comes from a sitting president, and unremarkable because it’s a single uncontrolled observation with no way to verify what information the psychic actually had access to.

Level 4: What It Means for Our Understanding of Consciousness

If even a small fraction of the reported effects are genuine, they would represent a fundamental challenge to our current models of consciousness and information processing. The implication is not that “psychics are real” but that consciousness might interact with the physical world in ways we don’t yet understand. This is a genuinely open scientific question, not a settled debate in either direction.

The neuroscience of consciousness is itself in its infancy. We do not have a complete theory of how subjective experience arises from neural activity. This is the “hard problem” identified by philosopher David Chalmers. If consciousness can interact with distant objects, it would suggest something about the nature of awareness that current neuroscience cannot explain. If consciousness cannot interact with distant objects, the consistent statistical anomalies in the psi literature still demand an explanation, because they appeared in peer-reviewed studies conducted by credentialed scientists.

Either way, the Stargate program left a question that has not been answered. It has been dismissed, celebrated, and ignored, but not resolved.

Level 5: What a Proper Investigation Would Require

If someone wanted to definitively answer the remote viewing question, the research protocol would need to include:

Large sample sizes: hundreds of trials, not dozens. The effect sizes reported are small enough that statistical power requires volume.
Pre-registered protocols: methods and analysis plans published before data collection, preventing post-hoc adjustments.
Independent replication: at least three labs, with no shared personnel or funding sources, producing convergent results.
A plausible falsification criterion: a defined result that would be accepted as evidence that remote viewing does not exist.
Blinding beyond the viewer: the experimenter, analyst, and scorer must all be blind to the target.
Elimination of sensory leakage: including electromagnetic shielding, distance requirements, and temporal separation between viewing and target selection.

This protocol would cost money, time, and institutional willingness to be associated with a topic the mainstream considers pseudoscience. No major funding agency has volunteered. The question remains open not because it is unanswerable, but because the political cost of answering it is higher than the scientific value of the answer.

—

What the Declassified Documents Actually Show

In January 2017, the CIA published approximately 12 million pages of declassified records online as part of the CREST (CIA Records Search Tool) archive. Among them were the Stargate Project files, spanning the full 23-year history of U.S. government remote viewing research.

These are not dramatic reading. The documents are mostly session transcripts, administrative memos, funding requests, and evaluation reports. They show a program that operated at the margins, producing occasional hits and frequent misses, sustained more by institutional momentum and Cold War fear than by consistent results.

What the documents reveal is a picture more complex than either proponents or skeptics claim. There are session transcripts where viewers described target sites with uncanny specificity. There are also transcripts where viewers described targets that had nothing to do with the actual location. The hit rate was never consistent enough to be operationally reliable, and the program’s own internal reviews acknowledged this repeatedly.

The declassification also reveals the program’s administrative instability. The project changed names at least five times as it moved between agencies (CIA, DIA, Army INSCOM). Each transition involved new oversight, new budgets, and new supervisors who had to be brought up to speed. This is not the environment in which rigorous, long-term scientific research thrives.

The documents also show that senior intelligence officials were consistently skeptical of the program’s value, even while funding it. The 1995 AIR evaluation was not the first negative review. It was simply the last one that mattered.

Video Resources

—

Sources & Further Reading

Study	Funding / COI	Venue	Tier
Targ, R. & Puthoff, H. (1974). “Information Transfer Under Conditions of Sensory Shielding.” Nature, 251, 602-607. DOI	SRI International / CIA contract. Researchers were employees of the funding agency.	Nature (top-tier)	🟡
Utts, J. (1996). “An Assessment of the Evidence for Psychic Functioning.” UCI page.	CIA/AIR contract. Independent evaluator.	AIR evaluation report (government contract)	🟡
Hyman, R. (1996). “Evaluation of the Military’s Twenty-Year Program on Psychic Spying.” Skeptical Inquirer, 20(2), 21-28.	CIA/AIR contract. Independent evaluator.	Skeptical Inquirer (peer-reviewed skeptical journal)	🟡
Cardeña, E. (2018). “The experimental evidence for parapsychological phenomena: A review.” American Psychologist, 73(5), 663-677. DOI | PubMed	No listed conflicts.	American Psychologist (top-tier)	🟢
CIA Declassified Archives, Stargate Project (2017). CREST Archive	U.S. Government (declassified). No COI. Operational records, not research.	Government archive	🟡
National Research Council (1988). Enhancing Human Performance: Issues, Theories, and Techniques. Chapter 9: Paranormal Phenomena. DOI	U.S. Government commission. No COI.	National Academies Press	🟢

—

—

FAQ

Q: Is remote viewing real?

The honest answer is: we don’t know for certain. There is a body of peer-reviewed experimental evidence showing statistically significant results. There are also legitimate methodological criticisms of that evidence. The effect sizes are small, replication is inconsistent, and no physical mechanism has been identified. The evidence is suggestive but not conclusive.

Q: Why did the CIA shut down the Stargate program?

The program was shut down in 1995 after an independent evaluation by the American Institutes for Research (AIR). Two evaluators (one sympathetic, one skeptical) agreed that the statistical effect, if real, was too small to be useful for intelligence operations. The AIR evaluators concluded that remote viewers “consistently failed to produce actionable intelligence information.” The program was not shut down because it was debunked. It was shut down because it wasn’t useful enough.

Q: Who was the best remote viewer?

Joseph McMoneagle is generally considered the most accomplished remote viewer in the program. A former U.S. Army Chief Warrant Officer and intelligence officer, McMoneagle participated in thousands of remote viewing sessions from the early 1970s through 1995. He was awarded the Legion of Merit for his intelligence contributions, which reportedly included remote viewing sessions that produced actionable intelligence.

Q: Is remote viewing different from psychic readings?

Yes. Remote viewing is a specific, protocol-driven research methodology. It uses structured procedures, double-blind conditions, and formal scoring systems. Psychic readings, as commonly understood, are informal, uncontrolled, and not subjected to systematic evaluation. The Stargate program was designed to make the research of clairvoyance more scientific, not to validate commercial psychic services.

Q: Can I learn remote viewing?

The laboratory studies do not address whether remote viewing is a learnable skill that can be transferred to the general population. The Stargate program used a small number of trained viewers. Some organizations offer remote viewing training, but there is no scientific evidence that such training produces results comparable to the laboratory studies.

Q: What about the skeptical position? Isn’t this all just cold reading and confirmation bias?

Some of the criticism is valid. The original SRI experiments had methodological issues that were addressed in later studies. Confirmation bias is a real concern in any research involving subjective judgments. But the skeptical position that “all the evidence is worthless” is not supported by the peer review process that accepted the studies, and the position that “the CIA proved psychic powers” is not supported by the agency’s own assessment of operational utility. The truth is more nuanced than either extreme.

Q: What about the Uri Geller connection? Doesn’t that discredit the whole program?

Geller’s early involvement with the SRI program is a legitimate concern. He was tested by Targ and Puthoff, reported results that were later called “fraudulent” by Ray Hyman, and the researchers lost their government contract as a result. But the program continued with different subjects and improved methodologies. Joseph McMoneagle, the program’s primary viewer, was a military intelligence officer, not a stage performer. The Uri Geller episode shows that the program had a credibility problem early on, but it doesn’t invalidate every experiment that came after.

Q: If the CIA had real psychic spies, why would they shut the program down?

Because “real” and “useful” are different things. A statistically significant effect in a laboratory setting does not automatically translate to a reliable intelligence tool. An intelligence officer needs specific, actionable information: coordinates, names, dates. Remote viewing produced vague impressions that might include a few sentences of accuracy buried in pages of noise. The effect was too small and too inconsistent for operational use.

Q: Are the CIA’s Stargate documents available to read?

Yes. In January 2017, the CIA published approximately 12 million pages of declassified records online, including the Stargate Project files. The documents are available at cia.gov/readingroom/collection/stargate. They include session transcripts, administrative memos, funding reports, and the 1995 AIR evaluation.

Related Research