Dreams and creativity: how your sleeping brain helps solve problems

How can sleep help you think differently?

During REM sleep, the stage most associated with vivid dreams, the brain does not simply switch off. Imaging studies show reduced activity in parts of the prefrontal cortex, which helps with planning and self-monitoring, while emotional and visual regions such as the amygdala stay very active. The usual filter that says "that makes no sense" is turned down.

At the same time, sleep reworks what you lived through during the day. Recent memories are replayed and linked to older ones. Researchers think this sorting may be where unexpected connections form: a work problem, a childhood beach and a song from the radio end up in the same scene.

This is a hypothesis about how creativity can emerge, not a promise. Most dreams are ordinary, and many good ideas vanish before breakfast. What studies do suggest is that two moments are worth watching: REM-rich dreams in the second half of the night, and the drowsy threshold just as you fall asleep, when brief images and loose thoughts drift by.

Famous ideas linked to dreams: what we actually know

These stories appear in almost every article on the subject. Some come from the creators themselves, others were told much later. It is worth knowing which is which.

August Kekulé and the benzene ring

The German chemist proposed a ring structure for benzene in 1865. Twenty-five years later, in a speech in 1890, he told how he had dozed by the fire and seen atoms twisting like snakes, one of which seized its own tail. Historians still debate how literally to take a story told so long after the fact. It remains a classic example of dream-driven discovery, with that caveat.

Paul McCartney and "Yesterday"

McCartney has often said that he woke up with the melody in his head, went to the piano and played it before it could fade. For weeks he asked people around him whether it was someone else's song, because it felt too complete to be his. It was his. The lyrics came later; at first he sang placeholder words.

Elias Howe and the sewing machine needle

The story goes that Howe dreamed of warriors whose spears had holes near the tip, and then placed the eye of his needle at the point. It is a good story, but no first-hand account from Howe is known, and other inventors had already tried needles with an eye near the point. Read it as a legend rather than a documented case.

Salvador Dalí and surrealist technique

Dalí described a way of harvesting the threshold of sleep in his 1948 book 50 Secrets of Magic Craftsmanship: doze in a chair holding a heavy key over a plate, and wake when it falls, with the last images still fresh. Thomas Edison is said to have napped holding metal balls for the same reason. A 2021 study tested a version of this trick, with striking results (see below). If you would rather stay aware inside the dream itself, that is the territory of lucid dreaming methods.

Mary Shelley and Frankenstein

In the introduction to the 1831 edition of Frankenstein, Mary Shelley recalls a night in the summer of 1816, near Lake Geneva. Lying in bed, eyes closed but unable to sleep, she saw a student kneeling beside the creature he had assembled. That waking vision, at the edge of sleep, became the seed of the novel.

What these stories share is not luck. Each person had been working hard on a problem or a project. The dream did not bring something out of nowhere; it rearranged material that was already there.

What research has measured

Laboratory studies cannot recreate the writing of a novel. They can, however, test whether sleep changes the way people solve problems with a hidden solution.

The Wagner study (2004): sleep inspires insight

In a study published in Nature, Ullrich Wagner and colleagues at the University of Lübeck trained participants on a number task that contained a hidden shortcut. When retested, more than twice as many people found the shortcut after a night of sleep as after the same time awake, whether that waking time was during the day or at night. The authors suggest that sleep restructures memories in a way that favours insight. The study did not record dreams, and it does not show which sleep stage was responsible.

Cai et al. (2009): REM sleep and remote associations

Denise Cai, Sara Mednick and their team used word puzzles in which you look for the link between three unrelated words. After an afternoon break, only participants whose nap included REM sleep clearly improved, compared with those who rested quietly or napped without REM. The authors conclude that REM helps combine information that was not previously associated. It was a small lab study, but a carefully controlled one.

Lacaux et al. (2021): the threshold of sleep

A team in Paris borrowed Edison's trick: participants rested holding a light bottle that would drop if they fell asleep. Those who spent at least 15 seconds in the lightest stage of sleep were nearly three times as likely to discover a hidden rule in a maths task afterwards (83% versus 30% of those who stayed awake). The advantage disappeared in those who slipped into deeper sleep. The sweet spot, in this study, is short and fragile.

What these studies do not show

These are simple tasks with a single right answer, tested on small groups. They show that sleep can help reorganise a problem you have already worked on; they do not show that a dream will write your next song. And they argue against one temptation: cutting your sleep to chase ideas. Sleep loss impairs attention, memory and creative thinking, so the healthiest strategy is still a full night.

How to take a question to bed

Dream incubation means choosing a question before sleep in the hope of dreaming about it. The idea is ancient: in Greek sanctuaries of Asclepius, visitors slept in the temple hoping for a healing dream. The modern version is far more modest.

In 1993, Harvard psychologist Deirdre Barrett asked 76 students to incubate a problem of their choice every night for a week. About half recalled a dream they judged related to their problem, and about a third of all participants felt the dream contained a solution. Personal problems were more often "solved" than academic ones, and independent judges were slightly more cautious than the dreamers. There was no control group, so this shows what is possible, not how often it works.

Step-by-step dream incubation technique

1. Write your question in one sentence. "How do I end chapter four?" works better than "I need inspiration".
2. Spend a few minutes with the problem in the evening. Look at your notes, sketches or drafts, then put them away.
3. Repeat the question as you fall asleep. Silently, or by picturing an image that sums it up.
4. Prepare to capture. A notebook or your phone within reach, so you can note the dream before getting up.
5. Look for images, not answers. A bridge, a door or a crowd can suggest a direction to test once you are awake.
6. Protect your sleep. No alarms during the night to "catch" dreams: use the awakenings that happen naturally.

For the full method and its variations, see our complete dream incubation guide.

How to keep a dream idea and use it

A dream idea fades quickly, especially if you move, talk or look at your screen first. Capturing it is mostly a matter of habit.

Note it before you move

On waking, stay still for a moment and replay the last scene. Then note everything, including fragments that seem absurd: the setting, the people, the dominant emotion. For techniques that strengthen recall over a few weeks, see our guide on how to remember your dreams.

Use your voice when you're half asleep

Writing is slow when your eyes are barely open. Speaking lets you keep up with the images before they dissolve. Tell the dream in the present tense, as if you were still in it: details come back more easily.

Reread with your project in mind

A dream idea often makes sense only later. Once a week, reread your notes next to your current work. Look for what comes back: a place, a colour, an emotion. Our dream journal guide explains how to build this habit.

Create a "dream idea bank"

Keep a separate list of images that could be useful, tagged by project. Add one line to each: what could I do with this? Most will stay curiosities. A few will become a sketch, a scene or a lead to test.

Example of a dream journal entry

Fictional example: it shows how to go from a dream image to an idea you can test, without asking the dream to decide for you.

• Dream: "I was in a library where the books floated like fish in an aquarium. Only one stayed still, open."
• Emotion: "Calm, curious. No hurry."
• Context: "For two weeks I've been stuck on a book cover that feels too crowded."
• Lead to test: "A single book, a lot of empty space, a deep blue. Sketch it this morning."

The dream did not solve the problem. It offered an image; the sketch will show whether it holds up.