Discovery Lab Answers

Dream Precognition: What the Evidence Shows

Coverage notice: The studies below reflect what the ESP-Nexus library currently holds on this question. The library’s share of the full published literature on dream precognition has not been measured, so this is a summary of what the library holds, not a settled account of the field. Treat the pattern described here accordingly.

What dream precognition is

Dream precognition is the claim that some dreams anticipate later events in ways not explainable by chance, inference, or memory. As Orme (2023) documents from historical case collections, the dream is consistently the most common reported vehicle for precognitive experience across multiple series of spontaneous cases — more common than waking visions or mediumistic states. The gap between how frequently people report the experience and how thin the controlled laboratory evidence remains is the central puzzle the research tries to close.

Laboratory approaches

The serious experimental work began at the Maimonides Dream Laboratory in Brooklyn in the 1960s–1970s. Post-Maimonides research has moved toward home-dream paradigms, free-response designs, and EEG presentiment protocols. Luke et al. (2012) investigated whether precognition performance fluctuates with circadian rhythms and melatonin cycles, training participants in dream-recall techniques and running sessions at different times of night; the results suggested some interaction between time of testing, dream recall, and psi scores, though few of the dream-recall trends reached significance, and length of dream recall appeared negatively related to precognition performance in that study.

Vernon, Roxburgh, and Schofield (2024) ran a home-dream paradigm in which participants rated images after waking, and the target image ratings were significantly higher than non-target ratings. The authors concluded the result was not attributable to methodological artifacts and described it as suggestive evidence for dream precognition, while calling for independent replication.

Robinson (2009) conducted a student-brief experimental study using a covert design and video targets, finding no evidence of participant preference bias or RNG selection bias in the data.

Quantitative evidence in the library

The structured evidence below spans a range of precognition paradigms — not all of them dream-specific, but relevant to the broader precognition question that dream studies sit within. The metric mix is important: hit rates, Cohen’s d, Cohen’s h, z-scores, and correlation coefficients appear across rows and cannot be averaged or pooled into a single bottom-line figure.

The unsettled signals — what is not resolved

Two features of this evidence base must be stated plainly:

Directional disagreement. Fifteen result rows across the structured evidence report null or below-chance outcomes. Several of these are large, well-powered studies — Walleczek (2025) with over 37,000 trials and 26,000 participants returned a null, as did the five-lab meta-analysis by Maier (2020) and all three of the Muhmenthaler (2022) replications of Bem’s retroactive priming paradigm. Positive results appear alongside these nulls, not instead of them.

Heterogeneity. The positive results themselves vary widely in effect size and paradigm. The very large z-values in Schmidt (2019) and Radin (2023) come from a handful of highly selected participants or from a specific card-position slope analysis across tens of millions of trials — neither generalizes straightforwardly to the dream context. Mueller (2019)’s associative remote viewing hit rate is strikingly high but comes from 15 participants over 48 trials. These are not comparable estimates of a single underlying effect.

Dream-specific coverage. Most of the structured evidence rows are precognition studies generally, not dream precognition studies specifically. The dream-specific controlled literature is smaller. Vernon et al. (2024) represent a recent home-dream paradigm with a positive finding, but themselves call for replication. Luke et al. (2012) found mixed circadian effects within dream sessions. Independent replication of home-dream paradigm findings outside the original research groups has not been published at scale.

Skeptical critiques

What critics argue. A recurrent methodological objection — visible in the Muhmenthaler (2022) replication program — is that positive results in precognition research, particularly those from Bem’s retroactive priming paradigm, do not survive pre-registered, adequately powered direct replication. Muhmenthaler and colleagues conducted three pre-registered replications of Bem’s Experiments 3, 4, and 9; all three returned null results with effect sizes near zero (d = 0.001, 0.034, and −0.03 respectively). Critics have also raised concerns about selective reporting, researcher degrees of freedom, and inadequate blinding in dream-recall paradigms, where judges who know the experimental hypothesis may unconsciously favor target-congruent matches.

What the experimental data show. The Bem (2016) meta-analysis, pooling 90 studies with over 12,000 participants, reports a pooled Hedges’ g of 0.09 in the positive direction. Mossbridge (2023) and Mossbridge (2024) report positive EEG presentiment results. At the same time, Walleczek (2025) — a large pre-registered study — returned p = .979 on erotic-trial precognition, and Schlitz (2021)’s confirmatory retroactive priming study returned d = 0.03, p = .22. Within dream research specifically, Vernon et al. (2024) report a significant result but acknowledge the need for replication.

Analysis. Pre-registration and direct replication have produced null results where earlier exploratory work was positive — this is documented in the Muhmenthaler (2022) rows above. The Bem (2016) pooled positive effect and the Muhmenthaler (2022) null replications of specific Bem protocols sit in the same evidence base, reported by different labs using overlapping methods and getting opposite outcomes. What specific procedural or population variables account for that difference has not been established in print.

For the site’s full synthesis on dream precognition specifically, including the Maimonides laboratory history and methodological discussion, see the Dream Precognition page. Simon J. Sherwood’s review of post-Maimonides dream ESP studies, covering meta-analytic findings and methodological advances, is available at https://esp-nexus.org/scientists/simon-j-sherwood/an-updated-review-of-dream-esp-studies-conducted-since-the/.