Research archive

The science is nineteen years old.
The access is new.

Targeted Memory Reactivation means replaying a study-linked cue while you sleep. It has been independently replicated since 2007. Every study below is mapped: effect sizes and null results included, not a cherry-picked headline.

54
Studies
1,300+
Participants
19
Years of research
The mechanism · How sleep consolidates memory

How TMR works.

During specific sleep phases your brain decides what to keep. Replaying a study-linked cue in that window reactivates the right memory, and scent wires straight to the hippocampus, so it cues strongly without disturbing sleep. On fMRI, the strength of that reactivation predicted next-morning recall (r = 0.70). Northwestern, 2018.

01 · Encode

Scent binds

You study with the cue present, binding it to the material. Sleep-time delivery reactivates that association.

02 · Consolidate

Memory replays

In deep sleep the cue returns and the hippocampus replays the linked memory, and your notes loop alongside.

03 · Transfer

Long-term storage

Slow oscillations and sleep spindles move the reactivated memory into cortex, automatically and below awareness.

04 · Retrieve

Improved recall

Re-smelling the cue before an exam reopens access to what was consolidated. This step shows the largest effects in the literature.

The evidence · 54 TMR studies

Every study, mapped.

◇ Interactive · maintained in code

Compare the approaches, then open any study. Effect sizes and null results are included. Filter by cue type, protocol, and outcome.

How long it lasts · Does the TMR effect fade?

Does the boost fade?

The benefit holds for at least a week, and in the scent studies, out to two weeks.

Memory naturally decays in days. The whole point of consolidating it during sleep is that it's still there when you sit the exam. Not gone by Thursday.

◇ Interactive · maintained in code

Each study's memory gain, placed by how long after learning it was tested. This is the question that actually matters for an exam, so it gets the last word.

Straight answers

Common questions about sleep, memory, and TMR.

Yes. In the foundational 2007 Rasch study, re-presenting a study scent during deep sleep raised memory for object locations to 97% versus 86% without the cue. A 2020 school study (Neumann) found students who had a scent present at studying, all night, and during the test scored about 82% versus 65% for a no-scent group. The archive shows all 54 studies, including mixed and null results.
Across the studies charted here, the benefit holds for at least a week after learning, and in the scent studies it was still measurable out to about two weeks. Some effects build overnight rather than appearing immediately.
It has been shown outside the lab. A 2022 at-home audio system (Whitmore) cut overnight forgetting of locations by about half, and a 2023 wearable that puffed a scent during learning and sleep (Amores) lifted next-day recall, with the gain lasting over a week.
Yes. The cue has to be tied to learning. In the 2007 Rasch study, a scent given during deep sleep that had not been present while studying produced no benefit. The cue during studying is what gives the sleep-time delivery something to reactivate.
Olfactory signals connect directly to the amygdala and hippocampus, largely bypassing the thalamus, so a scent can reactivate associated memories during deep sleep without disrupting it. Northwestern researchers captured this on fMRI in 2018: odor cues during deep sleep triggered category-specific reactivation, and the strength of that reactivation predicted next-morning recall (r = 0.70).
No. And the archive is deliberate about showing that. Several studies found the benefit depended on the person or the conditions, and on how well the material was learned going in. The case rests on the completeness of the evidence, not a single headline number.
About this evidence

This page maps every controlled targeted-memory-reactivation experiment we could locate: 54 studies and more than 1,300 participants, independently replicated since 2007, with null and mixed results included. Effect sizes labelled "computed" were derived by exact arithmetic from each paper's reported numbers; every study links to its primary source. Compiled and reviewed by the REMory research team. Last updated June 2026. This is an account of published research, not medical advice. REMory is a study aid, not a treatment.

Join Cohort 4

Nineteen years of research.
Now in your hands.

REMory is accepting applications for Cohort 4 from CPA, CFA, MCAT and LSAT candidates. Cohort 4 is limited to 25 spots.

Join the waitlist →