Dream Telepathy: What the Maimonides Research Actually Found
TL;DR - Key Takeaways
- Access modern tools like Hypnos to decode your subconscious
Dream Telepathy: What the Maimonides Research Actually Found
By Ron van Cann · June 2026 · 7 min read
In 1962, a small sleep laboratory opened at the Maimonides Medical Center in Brooklyn, New York. It was staffed by researchers who took an unusual question seriously: could information pass between minds during sleep?
Over the next decade, the Maimonides Dream Laboratory produced the most systematic, rigorously controlled experimental investigation of dream telepathy ever conducted. The results were provocative. They were also contested, partially replicated, and — depending on who you ask — either a strong signal or an artefact of methodology.
What is less often discussed is what the research actually showed, as opposed to what believers and sceptics claim it showed. This is an attempt to describe the work accurately.
The Setup: A Genuinely Careful Protocol
The principal investigators were Montague Ullman, a psychiatrist and dream researcher, and Stanley Krippner, a psychologist with broad research interests in consciousness. They were later joined by Charles Honorton, who would go on to develop the Ganzfeld paradigm.
The experimental protocol was designed to address the most obvious sources of error:
A "receiver" — the dream participant — slept in a soundproofed room with EEG electrodes attached, monitored by a technician who could track sleep stages. In a separate room, with no access to the receiver or the monitoring equipment, a "sender" opened a randomly selected envelope containing a target image (an art print) and spent the night concentrating on that image, attempting to mentally transmit it.
When the EEG showed the receiver entering REM sleep, the technician activated a buzzer that woke them after a few minutes, and asked them to report their dream content into a tape recorder. This process repeated throughout the night.
The next morning, the receiver was presented with a set of images — sometimes four, sometimes eight — including the actual target and a set of decoys. They were asked to rate how closely each image matched their dream content. Independently, judges who had not witnessed the night's events also received the dream transcripts and the image set, and rated the correspondence without knowing which image was the actual target.
Statistical significance was determined by comparing the ranked correspondence of target images to what would be expected by chance.
What the Data Showed
Across the formal experimental series, the Maimonides team reported significant correspondence between target images and dream content at above-chance levels.
Ullman, Krippner, and Honorton published a summary of eight formal experimental series in their 1973 book Dream Telepathy: Experiments in Nocturnal ESP. Across these series, the hit rate — the proportion of trials where the target ranked highest or in the top half — exceeded what chance would predict, and this excess was statistically significant when combined across studies.
Child (1985), a Yale psychologist, conducted a careful meta-analytic review of the Maimonides data and concluded that the combined results were statistically significant — the effect could not be attributed to chance at conventional significance levels. Child was not a parapsychology advocate; he was a mainstream academic psychologist who approached the review with scepticism and concluded the data warranted serious attention.
Some individual sessions produced striking correspondences. In one widely cited example, the target image was Salvador Dalí's The Sacrament of the Last Supper. The receiver's dream reports included references to a group of men, a setting that felt ceremonial or religious, a sense of being part of something larger. The imagery was not a perfect match, but the thematic correspondence was notable enough that both the receiver and the independent judges rated it as the closest match from the set — correctly.
The Criticisms: Where the Evidence Weakens
The Maimonides results have been criticised on several methodological grounds, and the criticisms are serious enough that treating the findings as established is not scientifically warranted.
The replication problem. The effect found at Maimonides proved extremely difficult to replicate in independent laboratories. The most direct attempted replication, by Belvedere and Foulkes (1971), conducted at the University of Wyoming using a comparable protocol, found no significant telepathic effect. Other replication attempts produced similarly null results.
Proponents argue that the Maimonides protocol required specific expertise in the sender-receiver relationship and in the experimental environment, and that naive replications inevitably miss important procedural details. This is a familiar argument in parapsychology, and it is also its central methodological vulnerability: an effect that only occurs under conditions specific to the original investigators is an effect that cannot be independently verified.
The judging problem. Even in the Maimonides studies, there were questions about how consistently the judging procedure was applied. The dream reports were sometimes long and varied; the matching task required judges to assess complex, ambiguous correspondence. Small differences in how instructions were given or how judges weighted different types of correspondence could influence the results.
The target selection and randomisation concerns. Several methodologists raised questions about whether the target selection procedure at Maimonides was truly random and whether target images were genuinely independent of each other. If certain target images (emotionally vivid, distinctive art prints) consistently produced more image-congruent dream content regardless of any telepathic effect, this could inflate apparent hit rates.
File drawer effects. The formal series published by the Maimonides team were selected from a larger body of trials. It is unclear how many informal or unreported sessions fell below significance — if these were more numerous, the published results would overrepresent positive outcomes.
The Ganzfeld Research: A Later Chapter
Charles Honorton left the Maimonides Laboratory and developed a different paradigm for testing anomalous information transfer: the Ganzfeld. In Ganzfeld studies, the receiver is not sleeping but is in a mild altered state (eyes covered, ears receiving white noise), while the sender attempts to transmit a randomly selected video clip.
Honorton and others conducted a substantial body of Ganzfeld research through the 1970s–1990s. A meta-analysis by Bem and Honorton (1994) reported a small but statistically significant hit rate above chance across the accumulated database. This analysis was published in the mainstream journal Psychological Bulletin, which was itself unusual for parapsychological research.
The Ganzfeld research was subject to the same replication critiques as the Maimonides work. Hyman (1985) conducted an independent meta-analysis of the same dataset and concluded that methodological flaws accounted for the apparent effect. A subsequent agreed-upon reanalysis by both Hyman and Honorton found continued disagreement about the methodology, though both conceded the results were not easily explained as chance.
The current scientific consensus is that the Ganzfeld and Maimonides dream telepathy findings do not meet the standard for established phenomena. No mechanism for information transfer between minds has been identified. Replication under fully blinded, independently conducted conditions has been inconsistent at best.
The Sheep-Goat Effect
One finding from parapsychological research that has received some mainstream attention is the sheep-goat effect: the observation that believers in psychic phenomena (sheep) tend to show significant above-chance performance in ESP tasks, while non-believers (goats) tend to show chance or below-chance performance.
Schmeidler and McConnell (1958) originally documented this effect in ESP card-guessing tasks, and it has been replicated across several subsequent studies. A meta-analysis by Lawrence (1993) found the sheep-goat difference to be statistically significant across a large combined dataset.
The effect is conventionally used by parapsychologists to argue that belief matters for the expression of psychic ability. The alternative mainstream explanation is that believers and sceptics differ in systematic ways in how they approach and respond to the task — with believers more likely to use certain response biases that happen to produce above-chance scores in typical experimental designs, and sceptics more likely to use opposing biases.
In dream research specifically, the sheep-goat dynamic is complicated by the fact that dreamers cannot control their dream content in real time. If the effect depends on motivated response strategies, it would be expected to be weaker in dream telepathy designs than in waking guessing tasks.
What Should a Careful Reader Conclude?
The Maimonides research was not pseudoscience. It was conducted by academically trained researchers, using a protocol that was more rigorous than most contemporary parapsychological work, and subjected to genuine statistical analysis. The positive results, combined with Child's independent meta-analysis, represent the strongest scientific case ever assembled for dream telepathy.
That case is not, however, sufficient to establish the phenomenon. The replication failures are significant. The absence of any identified mechanism matters. The methodological criticisms raised by Hyman, Belvedere, Foulkes, and others are not trivial.
The honest summary is this: the Maimonides experiments found something that deserves an explanation. The explanations on offer — genuine telepathy, or methodological artefact — remain contested. Independent high-quality replication has not produced consistent confirmation.
For people fascinated by the possibility of dream telepathy, the Maimonides work is genuine history and represents the high-water mark of scientific investigation into the question. For people interested in what science has actually established about the dreaming mind, dream telepathy remains, as of today, an open question at the margins of accepted neuroscience rather than a documented phenomenon within it.
Explore the contents of your own dreaming mind: Hypnos AI Dream Journal is free on the App Store.
Found this helpful?
Save this guide to your Dream Board.
