# Evening routine: what the night does with your last 10 minutes

**Canonical address:** https://optimistra.com/en/blog/evening-routine-before-sleep
**Authors:** `michel-wozniak`, https://optimistra.com/team/michel-wozniak; `nathalie-de-marce`, https://optimistra.com/team/nathalie-de-marce
**Category:** `Stress and Emotions`
**Published:** 2026-11-18
**Updated:** 2026-11-18
**Reading time:** 10 min, 2052 words

## In brief

During sleep, the brain replays and consolidates what the day left behind, and it is selective: what has been marked as useful for tomorrow gets through, the rest far less. The best-measured evening habit is not positive thinking. It is writing tomorrow’s to-do list, in detail.

---

You often hear that the last minutes before sleep “program” the night. It’s an appealing phrase, and it rests on something real: the brain doesn’t switch to standby when you fall asleep, it works. Between what research establishes and what the phrase suggests, though, the gap is wide. It is also interesting, because the real mechanism is both more modest and more usable.

Here is what the measurements show, the evening habit that has been tested in a sleep laboratory, and what screens actually do.

## What we mean here by an evening routine

An **evening routine** is what you do with the last minutes of the day, between the moment you stop your activities and the moment you fall asleep. For many people, that slot is filled by default: a phone, a news feed, a mental run-through of tomorrow. Nobody decided it, and that is exactly what makes it available: all it takes is to fill it differently.

The question here is narrower than “sleeping well.” It is this: **does what you put into those minutes change what the brain does next?** To answer it, you first need to know what the brain does while you sleep.

## The brain doesn’t rest, it files

This is the best-established point in the whole subject. The reference review, written by Björn Rasch and Jan Born and published in *Physiological Reviews* in 2013, describes a mechanism documented in animals and humans alike. During slow-wave sleep, memories recently encoded in a temporary store (the hippocampus, for declarative memory) are **repeatedly reactivated**, which gradually redistributes them to long-term storage, the neocortex.

In rodents, researchers can even watch the place cells of the hippocampus replay, during sleep, the firing sequences recorded during waking hours, at 10 to 20 times the original speed. According to the authors, this replay activity is strongest in the first 20 to 40 minutes of sleep or rest after learning.

In other words, something really does happen during the night, and it concerns the day that has just ended. On that point, the popular phrase isn’t wrong. It goes off track on what comes next.

## What the night sorts, and by what criterion

If the brain consolidated everything, the night would be a recorder. It is a sorter, and the criterion it sorts by is the most surprising part of this research.

In 2011, Ines Wilhelm’s team in Lübeck published a strikingly simple experiment in *The Journal of Neuroscience*. A total of 142 healthy adults learned pairs of words. Only afterward were some of them told that their recall would be tested 9 hours later; the others were told to expect an entirely different task. Then some slept, and the others did not.

The result: among those who slept, the participants who had been **told** retained 117.29% of what they knew at the start, against 105.23% for the others, a significant difference (p = 0.014). And the authors add the finding that is the heart of the matter: the performance of the uninformed group that slept was not significantly different from that of the uninformed groups that stayed awake.

Sleep did not help those who didn’t know they would need it. As far as that word list is concerned, they slept for nothing. The effect appeared again on a visuospatial task and on a motor task, and it came with an increase in slow-wave activity and in the number of sleep spindles in the informed participants.

Rasch and Born generalize this result in their review: the consolidation processes at work during sleep, they write, are driven by motivational factors and specifically strengthen the memories that are relevant to our goals.

**That is the defensible version of “programming your night.”** The criterion is not the intensity of what you felt, or the color of your thoughts as you fall asleep. It is **tagging**: what you have marked, one way or another, as useful for what comes next.

The word *program* is still too strong, because it assumes a machine and a command. The idea that you can steer the sorting does hold.

## The evening habit that has been measured

That leaves one question: how do you mark something as useful for tomorrow without a laboratory or an experimenter’s instructions? A team at Baylor University tested the most ordinary answer there is: write it down.

The study by Michael Scullin and his colleagues, published in the *Journal of Experimental Psychology: General* in 2018, is to our knowledge the only study of bedtime writing to use **polysomnography**, the standard measure of sleep onset: electrodes on the scalp rather than the sleeper’s own report.

Fifty-seven adults aged 18 to 30, with no known sleep disorder, spent one night in the laboratory. After the electrodes were fitted and before lights-out, each of them wrote for 5 minutes, with a pen. Half were asked to note **what they had to do** over the coming days; the other half, **what they had just completed** over the previous days.

Three results:

- The participants who wrote their list of upcoming tasks fell asleep significantly faster: t(55) = 2.32; p = 0.02; effect size d = 0.63, which is a clear effect.
- **The more detailed the list, the faster they fell asleep**: r(26) = −0.39; p = 0.04.
- In the group that wrote what it had completed, the trend ran the other way: the more detailed the list, the longer it took to fall asleep (r(27) = 0.34; p = 0.07, a trend, which the authors present as such).

The choice of control group deserves a mention, because it is what makes the study solid. Rather than compare writing with nothing, the researchers compared **two kinds of writing on the same subject**, which differed only in their direction in time. A “no writing” group would have left the door open to a simple placebo effect.

### What the authors themselves say about the limits

This is often the most instructive paragraph of a scientific paper, and this one is no exception.

The study covers **a single night**, the first one spent in the laboratory. The authors chose it deliberately, because it is a mild stressor and commonly serves as a model of transient insomnia. It involves **healthy young adults**, whose average time to fall asleep stayed under 30 minutes even in the control group. The sample, large enough for the experiment, is in the authors’ own judgment too small for the correlational analyses. And the mechanism itself remains unknown: how exactly bedtime writing affects the time it takes to fall asleep, they write, requires further investigation.

They conclude by calling for a randomized controlled trial, in people with and without sleep-onset insomnia, to find out whether the effect holds outside the laboratory and over several nights. **That trial has not yet been carried out.**

## What about screens?

This is the most repeated piece of advice in the field, and the study behind it is excellent, provided you read its protocol.

Published in *PNAS* in 2015 by the team of Anne-Marie Chang and Charles Czeisler, it compared reading on a light-emitting e-reader with reading a printed book. The results are clear: 25.65 minutes to fall asleep against 15.75 minutes (p = 0.009), melatonin secretion delayed by **more than an hour and a half**, less REM sleep (109 minutes against 121), and reduced alertness the next morning for several hours.

Now, the protocol. There were 12 participants. They read for **4 hours every evening**, from 6 p.m. to 10 p.m., for **five evenings in a row** in each condition. The e-reader was set to **maximum brightness** for the whole session, and the authors themselves acknowledge that this setting may overestimate the real effect. And the room was lit at about **3 lux**, which is almost dark.

These conditions are designed to maximize a circadian effect, and they succeed. They do not describe 10 minutes on a phone in bed. Over so short a time, the amount of light received bears no comparison, and the shift in the body clock shown here has no time to occur.

That does not mean a screen is neutral in the last 10 minutes. The reason changes, though: it is no longer a matter of light, it is a matter of **content**. A news feed or an inbox delivers a series of unsettled matters, in no order and with no end. That is exactly the material Scullin’s work identifies as costly, and exactly the opposite of a written list with an end to it.

## What the research does not establish

Three points, and they matter because you often see them stated as fact.

**That the night “thickens” the networks of feeling bad.** You commonly come across the idea that ruminating before sleep amounts to spending hours strengthening the circuits associated with negative emotions. Rasch and Born are explicit here: the findings on emotional memory during sleep are **contradictory**. Some studies show emotional memories being strengthened; others show that the recall of negative images is *less* affected by sleep deprivation, which suggests the opposite. And the authors conclude that it remains an open question whether the consolidation of emotional memories really differs, in kind, from that of neutral declarative memories. On a point where many articles take a firm position, the literature does not.

**That “thinking positive” before sleep changes the night.** None of the open-access sources we consulted measures this. What has been measured is the **direction in time** of what people write, upcoming rather than completed, and not its emotional tone.

**That dreams do the work.** A well-known study by Erin Wamsley and Robert Stickgold, published in *Current Biology* in 2010, showed that participants who dreamed of a virtual maze they had learned at midday improved 10 times more than the other sleepers. The result is spectacular, and the numbers are worth knowing: of the 50 participants who took a nap, **four** reported a dream related to the task. The authors also take care to rule out the causal reading: they do not claim that dream experiences cause memory consolidation during sleep. In their view, the dream and the improvement are two consequences of the same reactivation process. It is an indication, not a method.

## What you can do with this in practice

From everything above, three solid things remain, and one practical consequence.

The solid things: the brain replays and consolidates during sleep; it does so selectively, in favor of what has been marked as useful for what comes next; and writing down what remains to be done, in detail, leads to falling asleep faster than ruminating or going over what is already done.

The practical consequence is modest and takes 5 minutes: **put tomorrow on paper before the last minutes of the day**, with concrete tasks rather than intentions, then keep those last minutes free of anything that brings in new unsettled matters.

Try it over seven evenings, with a measure that looks not at the night but at the morning after: one line written on waking, the first thing that comes to mind, with no comment. It is the same discipline of observation as [naming a sensation before the emotion](/en/blog/naming-emotions-body-sensations): you note, you don’t interpret. After seven mornings, those seven lines tell you what your mind has been working on.

This topic connects with two others we have written about: [coherent breathing](/en/blog/coherent-breathing-five-seconds), one of the simplest ways to fill 10 minutes without asking anything of anyone, and the question of [how long it takes to form a habit](/en/blog/how-long-to-form-a-habit), which applies here as it does elsewhere. If the exercise leads you to notice physical tension in the evening, the article on [stress hormones](/en/blog/stress-hormones-cortisol-adrenaline) explains what the body does with its day, and the one on [acute and chronic stress](/en/blog/acute-and-chronic-stress-recovery) explains why the return to calm matters more than the height of the peaks.

**A necessary clarification.** Nothing above is a treatment for sleep problems, and neither is this article. Established difficulty falling asleep, lasting for weeks, calls for medical advice. It is worth having it looked at, because care is available for it, and the first option is not medication. A blog article is not part of that care.

*Article written by Michel Wozniak and Nathalie de Marcé, and optimized by artificial intelligence.*

---

## Frequently asked questions

### Do the last minutes before sleep really affect the night?

Yes, though not in the way it is usually said. It is established that during sleep the brain replays what it has recorded and transfers it to long-term memory. It is also established that it doesn’t do this for everything: in the study by Ines Wilhelm and her colleagues, only the participants who had been told they would be tested the next day benefited from their night. The others remembered as if they had stayed awake.

### Should you write a to-do list before bed?

It is the best-measured evening habit to date. In the trial by Michael Scullin and his colleagues, 57 people wrote for 5 minutes before lights-out, under polysomnography: those who noted what remained to be done fell asleep faster than those who noted what they had just completed. And the more detailed the list, the faster they fell asleep. The authors point out that the study covered a single night, in healthy young adults, and they themselves call for a clinical trial.

### Do screens really keep you from sleeping?

The most cited study deserves a close reading. Twelve participants read for 4 hours every evening, five evenings in a row, on an e-reader set to maximum brightness, in a room lit at 3 lux. The difference in time to fall asleep was about 10 minutes, and melatonin was delayed by more than an hour and a half. Those conditions have nothing to do with 10 minutes on a phone: over so short a time, the content matters more than the light.

### Does thinking positive thoughts before bed improve the night?

No open-access source measures this in the terms in which it is usually claimed. On the contrary, the review by Björn Rasch and Jan Born notes that the findings on emotional memory during sleep are contradictory, and that it remains an open question whether emotional memories are consolidated differently from neutral ones. What has been measured concerns the content of what people write, not its emotional tone.

### How long does it take to make a difference?

The studies cited here cover a single night in a laboratory. They say nothing about a repeated practice, and no one has measured what seven evenings in a row at home would produce. That is a reason to keep up the habit long enough to form your own view, with the caveat that a personal observation is not a measurement.

---

## Sources

- [Scullin, Krueger, Ballard, Pruett & Bliwise (2018), The Effects of Bedtime Writing on Difficulty Falling Asleep — A Polysomnographic Study Comparing To-Do Lists and Completed Activity Lists, Journal of Experimental Psychology General](https://pmc.ncbi.nlm.nih.gov/articles/PMC5758411/)
- [Wilhelm, Diekelmann, Molzow, Ayoub, Mölle & Born (2011), Sleep Selectively Enhances Memory Expected to Be of Future Relevance, The Journal of Neuroscience](https://pmc.ncbi.nlm.nih.gov/articles/PMC6623736/)
- [Rasch & Born (2013), About Sleep's Role in Memory, Physiological Reviews](https://pmc.ncbi.nlm.nih.gov/articles/PMC3768102/)
- [Chang, Aeschbach, Duffy & Czeisler (2015), Evening use of light-emitting eReaders negatively affects sleep, circadian timing, and next-morning alertness, Proceedings of the National Academy of Sciences](https://pmc.ncbi.nlm.nih.gov/articles/PMC4313820/)
- [Wamsley, Tucker, Payne, Benavides & Stickgold (2010), Dreaming of a Learning Task Is Associated with Enhanced Sleep-Dependent Memory Consolidation, Current Biology](https://pmc.ncbi.nlm.nih.gov/articles/PMC2869395/)
