In August 1973, Uri Geller took part in a series of drawing tests at Stanford Research Institute in Menlo Park, California. Russell Targ and Harold Puthoff interpreted the results as evidence of perception outside familiar sensory channels. Their work later appeared in Nature, but publication came with an unusually explicit editorial discussion of weaknesses and disagreement. The surviving documents tell a more complicated story than either a government certification of psychic powers or a single definitive exposure.
This case concerns the Menlo Park laboratory work and the controversy surrounding its publication in 1974. It does not treat every later remote-viewing operation as part of the same experiment. Different subjects, tasks and organisations became involved over the following decades. Keeping the 1973 sessions separate makes it possible to ask what Geller was actually asked to do, what the investigators recorded, and which conclusions their critics disputed.
The shielded room and the drawings
The report “Experiments — Uri Geller at SRI, August 4–11, 1973” describes sessions using a shielded room and, in some trials, distant targets. The researchers wanted repeatable picture tasks through which they could vary conditions. They expressed a strong positive conclusion about Geller’s apparent perceptual ability. That judgement belongs to the investigators who wrote the report; its later presence in released intelligence files does not turn it into an independent CIA experimental finding.
The day log contains successes, ambiguous correspondences and failures. For 4 August it records a firecracker target followed by drawings including a drum, and a grape target followed by a grape drawing. For 6 August, a rabbit target produced no corresponding drawing and Geller passed. On 7 August the tree and envelope targets also failed to produce matching drawings. These unsuccessful sessions belong in the chronology alongside the pictures that impressed the researchers.
The mixture is important when judging a collection of target-response pairs. A striking resemblance invites attention, but an assessment also needs the unsuccessful attempts, alternative drawings, spoken impressions and decisions about passing. Otherwise the visible record becomes a selection of favourable examples. The log supplies material for examining those questions; it does not remove the need for a clearly defined way of scoring a response.
The published experiment was not identical to the diary
Targ and Puthoff’s Nature paper, “Information transmission under conditions of sensory shielding”, appeared on 18 October 1974. It reports a defined drawing series of thirteen experiments across seven days and says none were omitted from the presented results. That description should be read on its own terms rather than silently equated with every activity in an eight-day laboratory log. A daily record and a published experimental series serve different purposes.
The paper describes target-selection procedures intended to prevent advance cueing, including drawings selected after isolation and material prepared independently. It also acknowledges that preliminary envelope tests could not distinguish possible perception of the picture from information associated with a knowledgeable experimenter. Geller was only one subject in the paper: other sections dealt with remote scenes and exploratory physiological work. Treating all its results as Geller’s performance would misdescribe the study.
Nature published its reservations too
Nature’s accompanying editorial, “Investigating the paranormal”, is central to understanding the publication. It reported a divided set of referee recommendations and criticised vagueness in procedure, target selection, safeguards and the mixture of different pilot investigations. The editors nevertheless considered publication worthwhile as an invitation to scrutiny. Their decision was an argument for examining controversial work, not an announcement that the extraordinary interpretation had been established.
The editorial also distinguished the perceptual claims from widely publicised metal-bending and other physical feats. It said the paper did not supply evidence for those powers. That boundary prevents an important error: acceptance of a report about attempted information transfer cannot automatically validate a separate claim about changing an object. The object, procedure and evidential question must be identified for each purported result.
The investigators answered the criticism
On 22 October 1974, Targ and Puthoff wrote to New Scientist editor Bernard Dixon. They agreed that their Nature paper did not report paranormal metal bending, while defending its perceptual conclusions. They rejected the suggestion that they had overlooked ordinary electronic communication, describing restrictions on people in the target area and silence about a target until a response had been collected. They also described a marked transparent die intended to rule out electronics inside it.
The letter disputed how visits by George Lawrence and Ray Hyman had been represented. According to Targ and Puthoff, the visitors had seen informal demonstrations and tried their own tests, rather than observed the researchers’ controlled series. This is their rebuttal, not a neutral adjudication of the encounter. It identifies a concrete disagreement about which session was being criticised and who had controlled its conditions.
A criticism and a demonstrated mechanism are different
The dispute leaves several questions that should not be collapsed into one. Were the reported drawings genuinely produced under the stated restrictions? Were the controls sufficiently described to reproduce? Could ordinary information have reached the participant? Were the matching and selection rules appropriate? A positive answer to one question does not settle the others. In particular, a proposed channel of leakage is not proof that it was used in a particular trial.
The same restraint applies to the investigators’ defence. Describing precautions is relevant, but an investigator’s belief that no leakage occurred does not by itself establish an unknown perceptual channel. Independent repetition under procedures specified in advance would test a different and stronger proposition than retrospective agreement about an impressive sketch. The scientific controversy concerned these practical questions of observation, control and evaluation, as well as competing beliefs about psychic phenomena.
What the intelligence paperwork establishes
A memorandum dated 30 August 1973 announced that Puthoff and Targ would present their final report on 10 September, including work involving Geller and Ingo Swann. It shows that the laboratory claims were being formally circulated for discussion. It does not show that the audience accepted every conclusion. An invitation, a contractor’s report, a journal editorial and an experimental record each establish different facts about the history.
This distinction becomes especially important with declassified material. A release stamp dates an administrative action, not the experiment. The authority holding a paper may not have written it, endorsed it or repeated its procedures. Here the report’s provenance helps document official interest, while the named experimenters remain responsible for the experimental interpretation. Publication and preservation make a claim available to examine; neither replaces examination.
The later remote-viewing review
A released American Institutes for Research draft dated 22 September 1995 reviewed a much broader remote-viewing programme. It separated laboratory research from intelligence applications and included differing expert assessments by Jessica Utts and Ray Hyman. Its discussion of operational usefulness should not be substituted for a re-analysis of Geller’s 1973 drawings. The report is later context, explicitly a draft in this released version, concerning other experiments and activities as well.
The enduring interest of the Menlo Park case lies in the accessible disagreement. The investigators’ daily log, their published account, Nature’s reservations and their subsequent rebuttal can be compared. Together they establish an important, contested investigation of claimed psychic perception. They do not justify merging favourable laboratory interpretations, press reports, metal-bending stories and later intelligence work into a single proven supernatural achievement.
Sources and further reading
SRI, “Experiments — Uri Geller at SRI, August 4–11, 1973”, released copy: the daily record, including unsuccessful targets. 30 August 1973 memorandum and attached SRI material: transcription with links to original scans.
Russell Targ and Harold Puthoff, “Information transmission under conditions of sensory shielding”, Nature 251 (1974), 602–607, DOI 10.1038/251602a0. The paper and the separate editorial “Investigating the paranormal” were checked in the released scan compilation, PDF pages 1–8. The editorial is the journal’s assessment, not part of the authors’ conclusions.
Targ and Puthoff’s October 1974 letters, especially the 22 October letter to Bernard Dixon: original scan checked; searchable transcription also available. These letters present the researchers’ side of the dispute.
American Institutes for Research, “An Evaluation of the Remote Viewing Program: Research and Operational Applications”, draft dated 22 September 1995: later programme-level context. Sources checked 5 November 2026.
Location and related reading
Map note: The pin is the US Census Bureau’s 2024 internal point for Menlo Park city. It indicates the broad locality, not a laboratory room, SRI entrance or public visitor access. The 2007 headquarters photograph illustrates institutional context.
Compare the importance of original experimental records in Crookes and D. D. Home, or explore Spiritual & Psychic and the Historical Cases directory.