The envelope sat on a table somewhere in a government facility. Inside it: a set of coordinates, or sometimes a photograph, or the name of a location known only to the person who sealed it. In another room, a subject sat with a pen and a blank sheet of paper and tried to describe what was there.
The sketches and session transcripts that came out of those rooms are now part of a declassified CIA collection. They are real documents. The program that produced them was funded by real agencies, overseen by real contractors and eventually reviewed by a real evaluation team. What that evaluation found is the center of this file.
Declassified records confirm that United States government agencies funded and managed remote-viewing research for roughly two decades. The 1995 review commissioned to assess the program concluded that it had not demonstrated reliable value for intelligence operations. That is the documented record. What happened between those two points—and what the documents do and do not prove—is worth opening carefully.
What the Documents Say
The CIA’s declassified collection on remote-viewing research, released in full to the National Archives in 2017, runs to thousands of pages. It covers a program that operated under several names—among them GONDOLA WISH, GRILL FLAME, CENTER LANE, SUN STREAK and, finally, STAR GATE—across portions of the 1970s, 1980s and into the mid-1990s.
The documents include administrative records, session transcripts, research contracts, evaluations from individual phases of the program and the final 1995 review prepared by the American Institutes for Research (AIR) at the request of the CIA. That review was authored by statistician Jessica Utts and psychologist Ray Hyman, who were given access to program records and asked to assess whether remote viewing had demonstrated operational utility.
Utts argued in her portion that the statistical results from controlled laboratory experiments exceeded what chance would predict, and that the effect warranted continued study. Hyman disagreed on methodology grounds, arguing the experimental controls were insufficient to rule out normal explanations and that the statistical argument was not as strong as claimed.
The AIR summary that framed both assessments reached a more direct operational conclusion: remote viewing had not been shown to provide intelligence of sufficient accuracy and reliability to justify continued use in intelligence operations. The CIA accepted that conclusion and the program was officially closed. Those documents are publicly accessible. The disagreement between the two reviewers is part of the record. So is the CIA’s final position.
Who Approved, Managed and Reviewed It
Funding and management of the program shifted across agencies and decades. Early research in the 1970s was conducted at Stanford Research Institute (SRI) under physicists Russell Targ and Hal Puthoff, with support from the CIA and later the Defense Intelligence Agency (DIA). The DIA eventually became the primary managing agency through much of the program’s operational period.
Congressional interest in remote viewing was partly driven by reports—difficult to verify in detail—that Soviet intelligence programs were investigating similar territory. The Cold War logic was straightforward: if the other side was investing in it, the cost of not investigating was potentially higher than the cost of investigating. That reasoning is documented in program histories and does not require any assumption about whether remote viewing worked.
Army remote viewer Joseph McMoneagle is among the named participants in the declassified record. McMoneagle was awarded the Legion of Merit in 1984, with a citation referencing his contribution to intelligence operations—a document that exists and has been publicly discussed. What that contribution consisted of, and how it was evaluated at the time, involves session records whose interpretation remains disputed.
The AIR review in 1995 drew on access to program records and was presented to Congress before the CIA announced closure. The review process itself—two named researchers reaching different conclusions, with an integrating summary—is part of the publicly available record and can be read directly.
What Was Actually Carried Out
Laboratory sessions were conducted at SRI and later at Science Applications International Corporation (SAIC), where research continued into the early 1990s. In those sessions, subjects were asked to describe distant or hidden targets using only coordinates, random numbers or sealed envelopes as prompts. Transcripts and sketches were then compared against actual targets by judges.
Operational requests—instances where intelligence agencies reportedly submitted real targets for remote viewers to attempt—are referenced in the record, though the results of specific operational sessions are harder to evaluate cleanly. Session outputs were described in general terms: a building with a certain shape, water nearby, a sense of underground activity. Matching those descriptions to targets involved subjective judgment, and the program’s critics consistently identified that subjectivity as a core methodological weakness.
The Mars session of 1984—in which a remote viewer was asked to describe a set of coordinates on Mars dating to approximately one million years ago—is part of the declassified collection and has attracted significant public attention since the documents became widely available. The session transcript exists. It describes ancient structures and inhabitants. What the transcript represents scientifically is a matter of interpretation; the AIR evaluation did not treat sessions of that kind as evidence of validated ability.
What was carried out across the program’s lifespan was a sustained research effort—laboratory studies, statistical analyses, operational trials—that produced a large paper record. That record shows research was done. It does not, by its existence alone, show that the ability being studied was real or reliable.
YOUTUBE EMBED 1 — PRIMARY SEARCH QUERY: “CIA Project Stargate declassified documentary”
The Missing, Limited or Disputed Record
The 1995 AIR evaluation is frequently cited in popular accounts of Project Stargate, but its internal disagreement is sometimes flattened. Utts’s statistical argument—that laboratory results exceeded chance—is cited by proponents of remote viewing as official acknowledgment that the effect is real. Hyman’s methodological objections and the AIR summary’s operational conclusion are cited less often in those same accounts.
The gap between laboratory results and operational reliability is the central tension in the record. A statistically above-chance result in a controlled experiment is not the same as a reliable intelligence source. The AIR review addressed that distinction directly. The program was not closed because the laboratory results were zero; it was closed because those results had not translated into actionable, accurate intelligence that could be independently verified and operationally trusted.
Some session records remain partially redacted or incomplete in the publicly released collection. The criteria used to match session outputs to targets—and who made those judgments, under what conditions—varied across phases of the program, making a clean aggregate evaluation difficult. Researchers who have studied the collection have noted these inconsistencies.
There is also the question of operational results that were never formally written up. Program participants have stated in interviews that some sessions were considered useful by handlers at the time. Those accounts exist and have been published. They are not, on their own, verifiable against independently confirmed intelligence outcomes.
YOUTUBE EMBED 2 — SECONDARY SEARCH QUERY: “CIA remote viewing archive explanation”
Claims the Evidence Does Not Support
The release of the CIA collection does not constitute scientific validation of remote viewing. The documents show the program existed, was funded and was reviewed. They record what participants produced. They do not independently verify that those outputs matched targets through any mechanism beyond normal sensory or analytical processes.
Claims that the program proved human paranormal ability are not supported by the AIR evaluation, which was the program’s own commissioned final review. Claims that specific operational successes—particular intelligence breakthroughs attributed to remote viewers—have been confirmed by the record are difficult to substantiate cleanly, because operational session results were evaluated subjectively and the independent confirmation trail is thin.
The Mars session is sometimes presented as evidence of extraordinary ability. The session transcript is real. Its contents are extraordinary claims. The AIR evaluation framework did not treat historical or speculative sessions as evidence of validated operational ability, and no independent evidence has established what the transcript describes.
The program also did not end because officials became convinced it worked and feared releasing that information. The documented reason for closure is the one in the AIR report: insufficient evidence of reliable operational value. That conclusion was presented to Congress and accepted by the CIA. It is in the public record.
Why This Story Still Survives After Midnight
The honest answer to the central question is this: the 1995 evaluation found that remote viewing had not demonstrated reliable value for intelligence operations, and the program was closed on that basis. The CIA’s own commissioned review, authored by named researchers with access to program records, reached that conclusion. It is documented, accessible and specific.
What keeps the file open is a real tension in that same record. The program ran for roughly twenty years. It received funding from multiple agencies. Named researchers produced statistical arguments that the two final reviewers did not fully agree on. Session transcripts and sketches are preserved in a government archive. That is an unusual amount of institutional attention for something that produced nothing at all.
The distinction worth preserving is between “the research existed and produced contested results” and “the research confirmed the ability worked reliably.” The first is documented. The second is not. The program’s history is a record of genuine institutional investment, genuine methodological disagreement and a final evaluation that came down on the side of insufficient evidence.
The sealed envelope on the table is a real object in the historical record. What it proves about what the person in the other room could perceive is still, by the standard of the program’s own final review, an open question the evidence did not close in the way some later accounts suggest.
