← all research

build dream recall

capture them before the memory fades

dreams recorded with a log versus without, across diary studies

what we know

act before the dream fades. the arousal-retrieval model, formalized in 1976, says a dream needs a brief waking window to become a memory. new sights, thoughts, and tasks can then interfere with it.

capture whatever remains before you do anything else. a scene, feeling, or single detail can be enough. in diary studies, keeping a log roughly doubled recorded recall. people who started with low recall gained the most.

attention also matters. in a study of 217 adults, a positive attitude toward dreams, more mind-wandering, and lighter sleep predicted higher recall. this means recall can rise with attention and awakenings. it does not show how well you slept.

the evidence

grade asystematic review, replicated controlled study, strong randomized trial, or objective physiological verificationstrongest supportgrade bgood controlled or longitudinal study, useful review, or converging evidence with meaningful limitationsuseful, with limitsgrade csmall, cross-sectional, exploratory, or under-replicated — a possibility, not a promiseearly or exploratorygrade danecdote, historical claim, or hypothesis — culture and questions, never proofstory, not evidence

dreams need an encoding window or they vanish

koulack & goodenough · psychological bulletin

1976grade bgood controlled or longitudinal study, useful review, or converging evidence with meaningful limitationssource ↗

interference after waking reduces recall

cohen · laboratory awakening experiments

1974grade bgood controlled or longitudinal study, useful review, or converging evidence with meaningful limitations

logs raise recorded recall, most for low recallers

aspy, delfabbro & proeve · methodological review

2015grade bgood controlled or longitudinal study, useful review, or converging evidence with meaningful limitationssource ↗

prospective recording beat memory-based estimates

zadra & robert · consciousness and cognition

2012grade bgood controlled or longitudinal study, useful review, or converging evidence with meaningful limitations

attitude, mind-wandering and light sleep predicted recall

elce et al. · communications psychology, n=217

2025grade bgood controlled or longitudinal study, useful review, or converging evidence with meaningful limitationssource ↗

immediate, neutral, verbatim capture is the least biased method

putois et al. · methods consensus

2020grade bgood controlled or longitudinal study, useful review, or converging evidence with meaningful limitations

“i dreamed, but it’s gone” is a real, common state

schredl · two-week diary study

2021grade csmall, cross-sectional, exploratory, or under-replicated — a possibility, not a promise

recall volume is not a marker of sleep quality

vallat et al. · polysomnography of high recallers

2022grade bgood controlled or longitudinal study, useful review, or converging evidence with meaningful limitations

use it in practice

totem asks one question before you stand up. capture the dream first, then start the day. fragments and feelings count. “i dreamed, but it’s gone” also records a real recall state.

totem tracks the capture habit. it does not use dream recall to score your sleep.

what we do not know

a journal can improve what you record and recall. this evidence does not show better sleep, health, or wellbeing.

guides built on this research

choose another question

remember more

try 7 nights free

7 nights free on the yearly plan · cancel any time