There are plenty of stories about remote viewing: the downed Soviet bomber, the CIA's secret programme, the silver futures on the commodities market. What usually gets lost in those stories is the least spectacular part of it – and at the same time the decisive one: the procedure. Remote viewing is less a gift than a set of rules. Taken seriously, it works with target pools, blinded participants, standardised stages and an evaluation the viewer cannot influence at all. This article explains how a session actually runs, how it is scored, what the aggregated evidence shows – and where the criticism bites that you need to know about.
What remote viewing is – and is not
Remote viewing means attempting to obtain information about a place, a person or an event to which the describer has no sensory access. The term was coined at the Stanford Research Institute (SRI) in the 1970s, and it was deliberately chosen to sound dry: the researchers wanted an expression that suggested neither occultism nor worldview, but a task in a laboratory.
The difference from neighbouring terms lies less in the experience than in the surroundings. Clairvoyance describes an ability; remote viewing describes an ability under protocol. And unlike mediumship, it is not about contact with the deceased but about a factual description of a target – although some practitioners do both. The portrait of Birgit Fischer shows that double role from the inside.
It also matters what remote viewing does not claim. It says nothing about how the information arrives. Whether one regards consciousness as non-local or suspects some unknown physical coupling – the protocol is independent of that question. It only measures whether descriptions match their targets better than chance allows.
The core: the protocol, not the gift
The real achievement of the SRI physicists Harold Puthoff and Russell Targ was not finding extraordinary people but building an arrangement in which nobody can cheat by accident. Four components carry the procedure:
- The target pool. Before the session a set of possible targets is fixed – say a hundred locations in the surrounding area. For each trial someone draws a target at random. Only this makes the later arithmetic meaningful: you know how many alternatives the description is competing against.
- Blinding. The viewer does not know the target. In the strict case (double-blind) neither does the person sitting in the room running the session – otherwise they may give it away unintentionally through tone, follow-up questions or body language. This is precisely where almost every serious criticism starts.
- Access to the target. Historically there were two routes. In the outbounder or beacon method a staff member travels to the drawn location and stays there while the viewer describes it. Later came coordinate remote viewing, in which the viewer receives only geographical coordinates or even just a meaningless random number. The number is the more radical approach: it demonstrably carries no information about the target.
- Feedback. After the session the viewer learns what the target was, usually through a site visit or a photograph. Practitioners consider this necessary for learning. Methodologically it is delicate, because feedback from earlier trials can influence later ones; this is why target pools are normally not reused between trials.
A fifth term comes up constantly in practice: frontloading. This means any advance information the viewer receives about the target – for instance "it is a building". In operational use a little frontloading is sometimes unavoidable. For evidential purposes it is poison, because hits can no longer be cleanly separated from prior knowledge.
The six stages of coordinate remote viewing
Between the mid-1970s and the mid-1980s Ingo Swann led the development of a teachable methodology at SRI. The result was coordinate remote viewing (CRV), later often called controlled remote viewing. The underlying idea: the problem in describing an unknown target is not a lack of impressions but premature interpretation. Whoever turns a line into "bridge" straight away has already lost the session. The stages therefore force the viewer to record raw perception first and to postpone meaning as long as possible.
| Stage | What happens there |
|---|---|
| I | Ideogram – a spontaneous pen movement towards the target, plus the rough gestalt (land, water, structure, motion). |
| II | Sensory qualities: colours, temperatures, textures, sounds, smells – noted individually, without summarising. |
| III | Dimensions and sketches: tall, wide, deep, open, closed – the first spatial impression. |
| IV | Ordering the impressions in a matrix, including the explicit separation of perception from the viewer's own interpretation. |
| V | Targeted interrogation of individual impressions in order to obtain analytical statements about the target. |
| VI | Three-dimensional modelling – the viewer builds or sketches the target and how its parts relate to one another. |
In 1983 and 1984 six members of the military unit at Fort Meade were trained in this procedure; the training manual Tom McNear wrote from it in 1985 reads like a workbook with session transcripts and marginal notes. An important qualification: the stages are a discipline technique, not a proof of efficacy. Whether CRV yields better results than free description has never been cleanly tested head to head.
Scoring: why the ranking is the actual trick
Here is where remote viewing differs from any living-room demonstration. A description such as "something round, metal, water nearby" always sounds somehow apt once you know the target. Which is exactly why nobody who knows the target may do the judging.
The standard procedure is rank-order judging. An independent judge who knows neither the session nor the assignment receives the transcript and all the targets from the pool – five locations, say. They visit them or view photographs and put them in order: which location fits the description best, which second best? If the first rank is correct more often than would be expected with five alternatives, that is statistically tractable. The viewer cannot influence the result, and the judge cannot drift towards a target they do not know.
The most famous early run of this kind involved Pat Price. Five SRI scientists otherwise uninvolved visited nine locations independently and matched the transcripts to them; in six of the nine cases the majority assignment was correct. This was published in Nature in 1974 – the reason the topic reached mainstream science at all.
Besides ranking there is the rating of individual statements and sketches on a scale, often additionally by the viewer. This yields finer-grained values but is more open to interpretation – and is therefore usually reported only as a supplement.
Associative remote viewing: the branch for yes-or-no questions
The protocol can only describe, not answer. Ask a viewer directly whether the silver price will rise and you no longer have a target pool or a clean evaluation – and field experience is that direct yes-or-no questions barely work.
The workaround is associative remote viewing (ARV). Each possible outcome is linked in advance to an object: if the price rises, the viewer is later shown a pine cone; if it falls, a seashell. What gets described is therefore not the market but the future feedback object. What this looks like in practice – and why it failed when scaled up – is covered in detail in the article on Delphi Associates and the 1982 silver futures.
What the evidence shows
Since 2023 there has been an aggregate evaluation of the whole field for the first time. Patrizio Tressoldi and Debra Katz reviewed every remote viewing study traceable up to the end of 2022 and, after applying inclusion criteria, analysed 36 studies with 40 effect sizes. The result after excluding outliers: an average effect size of 0.34 (95% confidence interval 0.22–0.45), with no sign of publication bias and only a slight decline over time. Translated into raw scores, that corresponds to a hit rate 19.3 percent above chance expectation (confidence interval 13.6–25 percent).
That number needs interpreting. It does not say that a viewer "sees" a target – it says that across five decades and several laboratories the assignments come out systematically better than expected. Tressoldi and Katz conclude that the remote viewing protocol performs above average compared with other experimental designs for extrasensory perception. The same researcher, incidentally, also produced the meta-analysis of mediumship research.
The second reference point is older and had greater consequences. When the CIA had the government programme evaluated in 1995, the American Institutes for Research commissioned two assessors: the statistician Jessica Utts and the psychologist Ray Hyman. Utts concluded that by the standards applied in other scientific fields the effect was sufficiently established. Hyman disputed this. What is remarkable is where the two agreed: that the effect sizes of the later studies were consistent, and that the results could not be explained away by the usual standard methodological objections. Hyman's objection amounted to this: a statistically stable effect without a mechanism and without independent replication outside the field is not yet a scientific fact. This debate is set out in detail in the article on the Stargate programme.
The other side, taken seriously
Anyone interested in remote viewing should know the three strongest objections – not as an obligatory exercise, but because they sharpen the sense of what a clean protocol has to achieve.
Cues in the transcripts (1978). The psychologists David Marks and Richard Kammann tried to repeat the SRI results and found something awkward: the published transcripts contained asides from which the order of the locations visited could be inferred – dates, references to earlier trials. A judge could use these to assign correctly without any perception being involved. Marks and Kammann published this in Nature in 1978, and Marks followed up in 1981. The other side (Tart, Puthoff and Targ) had the transcripts cleaned and re-judged and reported that the result held; the critics in turn disputed the care taken over that correction. The dispute was never fully resolved. Its lasting yield is a rule that is taken for granted today: transcripts going to judges must be stripped of every piece of contextual information.
Information leakage in the successor programme (1998).Richard Wiseman and Julie Milton examined the first experiment of the SAIC programme – the very studies Utts and Hyman regarded as the cleanest – and described possible routes by which information could nonetheless have reached the judge. Edwin May objected; Wiseman and Milton replied. Reading the exchange, one quickly notices that it is not about fraud but about how many improbable channels must be excluded before a result counts.
The gap between statistics and application. The most honest objection comes from inside the programme. An effect of roughly 19 percent above chance is scientifically fascinating and operationally difficult all the same: if a description can be right or wrong in the individual case and you cannot tell in advance which, it is no basis for a decision on its own. That was the practical reason for the shutdown in 1995 – not proof that there was nothing there. Nor do the spectacular individual cases change this, such as those of Joseph McMoneagle, because individual cases cannot be blinded retrospectively.
What remains
Remote viewing is probably the best-documented attempt to translate a contested phenomenon into a testable procedure. Its yield is a methodology that many other areas of consciousness research have borrowed: target pool, blinded participants, independent judging, evaluation by ranking. Its second yield is a number that is small but stable across five decades.
What is missing is a mechanism – and the ability to say in the individual case whether a description is correct. Neither gap is trivial. But this is a different situation from the one commonly assumed: not a refuted field, but a finding without a theory.
Sources:
• R. Targ & H. Puthoff, Information transmission under conditions of sensory shielding, Nature 251, 602–607 (1974).
• H. Puthoff & R. Targ, A Perceptual Channel for Information Transfer over Kilometer Distances: Historical Perspective and Recent Research, Proceedings of the IEEE 64(3), 329–354 (1976).
• D. Marks & R. Kammann, Information transmission in remote viewing experiments, Nature 274, 680–681 (1978); D. Marks, Sensory cues invalidate remote viewing experiments, Nature 292 (1981).
• R. Wiseman & J. Milton, Experiment One of the SAIC Remote Viewing Program: A Critical Re-evaluation, Journal of Parapsychology 62(4), 297–308 (1998).
• J. Utts, An Assessment of the Evidence for Psychic Functioning, report for the American Institutes for Research (1995).
• P. E. Tressoldi & D. Katz, Remote Viewing: A 1974–2022 Systematic Review and Meta-Analysis, Journal of Scientific Exploration 37(3), 467–489 (2023).
• Coordinate Remote Viewing Training Manual (version following the SRI training, T. McNear 1985).
The history behind the method is told in our articles on the CIA Stargate programme and on the two SRI physicists Harold Puthoff and Russell Targ. There are portraits of the individual viewers Ingo Swann, Pat Price and Joseph McMoneagle. Applied cases are covered in the Delphi Associates silver futures, Martin Zoller's Bolivia operation (see also his portrait) and Birgit Fischer on remote viewing and the CIA. How things went in other countries is covered under worldwide psi programmes.