Everyone has tried it at least once. You lie down thinking very hard about a person, a place, a problem at work, and you ask the night to take it from there. Usually the night ignores you. Sometimes, unsettlingly, it does not.
So can you choose what you dream about? Not the way you choose a film. But you can tilt the odds, and in 2026 a sleep laboratory measured by how much. A team at Northwestern University steered sleepers’ dreams toward specific puzzles they had failed to solve before bed, and the puzzles that made it into a dream were solved more than twice as often the next morning. [1]
This post covers what they did, what the older research on “dream incubation” says, and a method you can run at home with nothing but a journal and a question. It ends with the honest part: what tends to come back is rarely an answer. It is usually something stranger and, with practice, more useful.
What the Northwestern team did
The study, led by Karen Konkoly and Ken Paller and published in Neuroscience of Consciousness in February 2026, recruited 20 adults, most of them frequent lucid dreamers, and brought each of them into the lab for two nights. [1]
On each night, before sleep, participants worked through a series of brain-teasers: matchstick puzzles, rebuses, word and spatial problems of the kind that need a sudden change of angle rather than patient arithmetic. They had three minutes per puzzle, and they kept going until four puzzles had beaten them. Every puzzle had its own fifteen-second soundtrack, a snippet of music or ambient sound that played on repeat while they worked on it. Before bed, they rehearsed the pairings until they could name the puzzle from the sound alone.
Then they slept, wired to a polysomnograph. At four in the morning the experimenters woke them, taught them a short cue meant to trigger lucidity, and let them go back to sleep with one instruction: if you hear a puzzle’s soundtrack in a dream, work on that puzzle in the dream.
When each participant entered REM sleep, the stage where the long, vivid dreams live, the researchers played the soundtracks of two of the four unsolved puzzles, quietly, over and over. The other two puzzles got no cues. This is a technique called targeted memory reactivation, and it is the cleverest part of the design: every sleeper was their own control group. Any difference between cued and uncued puzzles could not be down to who slept better or who was cleverer.
After each cued REM period, the sleepers were woken and asked to report their dreams. In the morning they got four more minutes on every puzzle.
What they found
Three findings matter for anyone who wants to dream about something on purpose. [1]
The cues worked. Puzzles whose soundtracks were played during REM turned up in dreams far more often than puzzles left alone. Fifteen of the twenty participants had at least one dream about an unsolved puzzle, and on average people dreamed about a quarter of the puzzles they had failed. In some dreams the puzzle appeared outright. In others it arrived sideways:
I was at a river and I was catching fish with a net. I was working on something, like a puzzle. So it could have been the jungle one.
That participant solved the jungle puzzle the next morning.
Dreaming about a puzzle predicted solving it. Across all puzzles, the ones that had appeared in a dream were solved the next morning far more often than the ones that had not.
| Puzzles that | Solved next morning |
|---|---|
| Appeared in a dream | 42% |
| Did not appear in a dream | 17% |
Merely hearing the soundtrack in a dream, without the puzzle itself showing up, did nothing. The benefit was tied to the content, not the sound.
For the people whose dreams took the cue, the cue improved solving. Twelve of the twenty participants were what the authors call targeted dreamers, meaning cues reliably pushed their dreams toward the cued puzzles. For that group, cued puzzles were solved 40% of the time and uncued puzzles 20%. For the other eight, whose dreams ignored the cues, cueing made no difference at all. That is the pattern you would expect if the dream itself is doing the work.
Two smaller results are worth keeping. One is that the effect carried on after the lab: in the week of home dream journaling that followed, about one dream in eight still referenced an unsolved puzzle, and the trend ran the same way. [1] The other is about lucidity, and it is surprising enough to get its own section below.
The catch, which is also the point
The authors are careful, and so should we be. The sample was twenty people, most of them selected for lucid dreaming, and the study was underpowered for several of the comparisons it ran. Participants knew the study was about dreaming and puzzles, so expectation could have played a part. The most important limit is this: everyone was told to dream about the puzzles before they went to sleep. The researchers cannot separate the effect of the sound cues from the effect of that pre-sleep intention. [1]
For a scientist that is a confound. For you it is the useful part, because you do not own a sleep lab and cannot play soundtracks to yourself during REM. What you can do is form an intention before sleep. And that, it turns out, has a research history of its own.
Dream incubation, the old technology
Asking the night a question is an ancient practice. Pilgrims slept in the temples of Asclepius hoping to dream a cure, and when people are surveyed today, a steady minority credit a dream with a creative idea they later used. [9] The modern name for doing it on purpose is dream incubation, and there is a small but consistent literature on whether it works.
In 1993 the Harvard psychologist Deirdre Barrett asked 76 students to pick a real problem, one with personal or academic stakes, and to spend a few minutes each night for a week focusing on it as they fell asleep, with a notepad by the bed. Roughly half reported a dream that related to the problem within the week, and about a quarter reported a dream that, in their own judgement and that of independent raters, contained a solution. [4] A later study compared incubating a personal problem before sleep with simply relaxing before sleep, and found that the incubators rated their problems as more solvable afterwards. [5]
The research on sleep and insight more broadly backs this up. In a well-known 2004 experiment, people who slept after practising a tedious number task were more than twice as likely to discover a hidden shortcut the next day as people who stayed awake. [6] In 2021, a French team found that a single minute spent drifting into the first stage of sleep, and then waking, nearly tripled the chance of spotting the same kind of hidden rule. [7] Sleep onset is a creative sweet spot, as the title of that paper put it.
The Northwestern group had already shown in 2019 that playing puzzle soundtracks during deep sleep at home, with a wearable headband, raised next-day solving from 21% to 32%. [2] And a team at MIT built a device called Dormio that whispers a word to you as you drift off, then wakes you to ask what you were dreaming; people steered toward dreaming about a tree did better afterwards on creative tasks about trees than people who merely thought about trees while awake. [3]
Put the pieces together and the picture is consistent. Intention before sleep raises the chance of dreaming about a thing. Dreaming about a thing raises the chance of a new angle on it. Neither effect is huge. Both are real.
How to dream about something specific: a five-step method
This is the version of dream incubation that the research supports, adapted for a person with a journal and no electrodes. Run it for a week before judging it.
1. Choose one question, not a wish
The experiments that worked all used a single, concrete, unsolved problem that the dreamer actually cared about. “I want to dream about the sea” is a wish. “What should I say to my sister about the house?” is a question. Questions give the dreaming mind something to work on, and a puzzle you have already tried and failed is the ideal material, because the failed attempt is what the night has to rearrange. [1] [6]
One question at a time. Studies that gave people ten problems to sleep on found no sleep benefit at all, probably because nothing got incubated properly. [1]
2. Write it down, as the last thing you do
Phrase the question in one plain sentence and write it in your journal as the final entry of the day. Barrett’s students kept the problem by the bed and reread it; the Northwestern participants rehearsed the puzzle-sound pairings right before sleep. The common factor is that the problem is the freshest thing in memory at sleep onset.
Tonight: why does the opening of chapter three feel wrong?
Tonight: what am I actually afraid of about the move?
In Lunar Log, write the question as a short entry tonight; tomorrow’s dream lands beside it, and reading the two together in the morning is half the method.
3. Give the question a shape
The lab used soundtracks because a sound can be played to a sleeper. At home you can borrow the principle without the equipment. Spend a minute picturing the problem as a scene: the two people in the room, the page with the wrong paragraph, the locked door. If the problem has an object, put the object where you will see it last, on the bedside table. Barrett has argued that dreams favour problems that can be pictured, and in her own study personal problems were dreamed about and solved more often than academic ones. [4] [10]
4. Fall asleep quickly, holding the question lightly
Here the lucid dreaming research has a lesson. In the largest trials of MILD, the intention-setting technique that lucid dreamers use, success depended heavily on falling asleep within a few minutes of setting the intention; our lucid dreaming guide covers the details. The same logic applies here. Do the writing and picturing with the light low, lying in your sleeping position, phone elsewhere. Do not grind on the problem. The Northwestern participants were told to stop working on the puzzles before bed, and the whole point of sleeping on something is that effortful thought gets out of the way. [1] [6]
5. Record first, judge later
On waking, before moving much, record whatever is there, even a fragment, even a mood. Our guide to remembering your dreams explains why the first minute matters so much. Speak it if you can; voice capture is faster than typing with your eyes half closed. Do not ask yet whether the dream answered the question. Just get it down.
Then, over coffee, read last night’s question and this morning’s dream side by side. Keeping a logbook roughly doubles how many dreams you remember within a fortnight, so the method gets easier the longer you run it. [11]
What usually comes back
This is where most guides to dreaming about something specific go quiet, because the honest answer is: not what you asked for.
In the Northwestern study, the dreams that helped were almost never the puzzle laid out neatly with its solution. They were a river and a net. They were a friend’s daughter on a seesaw, which is how one dreamer noticed the scales in her puzzle were not level. One participant signalled from inside her dream that she was working on two cued puzzles, using the sniff-and-eye-movement code this lab developed for talking to sleepers in real time, [8] then remembered only “looking through a dresser” and “a little celebration” when she woke, and then solved both puzzles in the morning. [1]
Carl Jung would not have been surprised. He argued that a dream does not deliver a message in code; it shows you your situation from an angle you were not using, and the strangeness is the point. For how that works in practice, see What Do Dreams Mean? The dream you get after asking about your sister might be about a bridge, or a house with a room you had forgotten, or an old school friend. Write it down anyway. Then ask the five questions that post describes, starting with the feeling the dream left you with, which is usually the most direct link back to the question. Why Dreams Feel So Emotional makes the case for reading the feeling first.
Three things to look for when you compare the question to the dream:
- A change of frame. The problem appears, but the setting, scale or role has shifted. You are the sister. The house is a boat. That shift is frequently the useful content.
- An attempt. In the lab, dreams in which the dreamer actively tried something, asked a character for help, looked for the puzzle, were solved more often than dreams where the puzzle simply floated past. [1] If you find yourself doing something about the problem in a dream, note what.
- An emotion that does not match. You asked about a work decision and dreamed something tender, or something full of dread. The mismatch is information.
Why you do not need to be lucid
The Northwestern team recruited lucid dreamers on purpose, expecting that people who could knowingly work on a puzzle inside a dream would do best. The opposite happened. Of nine puzzles worked on during lucid dreams, one was solved the next morning. Of thirteen that appeared in ordinary, non-lucid dreams, six were solved. [1]
The authors offer a sensible guess. Deliberate effort inside a lucid dream may drag the mind back onto the same wrong track it was stuck on while awake, whereas an ordinary dream lets the wrong answers fade and the odd associations surface. Several of their lucid dreamers were also plainly sceptical of the instruction to ask the dream for help, and the scepticism showed up in the dreams:
I say, “How do you do the four-shape puzzle?” She’s like, “Well I actually don’t know. It’s kind of hard.”
So if you have never had a lucid dream, you are not at a disadvantage here. Intention before sleep and a journal in the morning are the two levers, and both are available to everyone. If you do become lucid and want to use it, the best-attested approach in the anecdotal literature is to ask the dream a question and then wait and watch, rather than to reason your way through. [1]
Common problems
I dreamed about it, but there was no answer
Expected. A quarter of Barrett’s students got a solution in a week; the rest got related dreams, or nothing. [4] The dream’s job is to move the problem, not to finish it. Keep the question for another night or two, and on the third morning reread all three dreams together. The pattern across them is often clearer than any one of them.
I dreamed about something else entirely
Also expected, and not a failure. Dream content tracks what is actually pressing on you, not necessarily what you nominated. [12] If you asked about the book and dreamed about your father three nights running, the night may be telling you which question it thinks you should be asking. Recurring content is worth its own treatment; see Why Recurring Dreams Keep Coming Back.
I remembered nothing
Then the question is not incubation but recall, and recall responds to practice faster than almost anything else in dreaming. Give it two weeks with the journal before you judge the method. [11] The night may well be working on your question; you are just not catching the report.
Can I make myself dream about a specific person?
The same method applies, with one adjustment: ask a question about the person rather than wishing for the person. “I want to dream about him” gives the dreaming mind nothing to do. “What is unfinished between us?” gives it a job. Be ready for the answer to arrive as someone else entirely, since people in dreams tend to stand in for something, as our page on dreaming about an ex explains.
Where a journal fits
Every study above had a journal in it. Barrett’s students kept a notepad by the bed. The Northwestern participants filled in a dream diary for a week after each lab night. The method does not work without the morning record, because a dream you do not write down cannot be compared to the question you asked.
Over weeks, the journal starts to show you your own incubation style: whether your answers come as scenes or as feelings, whether they come the first night or the third, which kinds of question your dreams take up and which they ignore. Lunar Log tracks the people, places and symbols that recur, so when the same figure turns up every time you ask about work, you will see it. That census is also exactly what you need if you later want to try lucid dreaming, where the whole method is built on knowing your own dreams well.
Frequently asked questions
Can you control what you dream about? Not directly, but you can bias it. In a 2026 study, playing sound cues linked to specific puzzles during REM sleep made people dream about those puzzles more often, and setting an intention before sleep has a similar, smaller effect in earlier research. [1] [4] [3] Nobody can guarantee a given dream on a given night.
How do you dream about something specific? Pick one concrete question, write it down as the last thing before sleep, picture it as a scene, fall asleep quickly without grinding on it, and record whatever you remember first thing in the morning. Expect the dream to approach the question sideways.
Does sleeping on a problem actually help? Yes, in controlled studies. Sleep after practising a task more than doubled the chance of spotting a hidden shortcut, and dreaming specifically about an unsolved puzzle raised the chance of solving it from 17% to 42%. [6] [1]
Do I need to lucid dream for this to work? No. In the 2026 study, puzzles that appeared in ordinary dreams were solved more often than puzzles worked on in lucid dreams. [1]
How long does dream incubation take to work? In the best-known study, about half of people had a relevant dream within a week of nightly practice. [4] Give it at least a week, and keep the journal going; recall improves over the same period. [11]
What is targeted memory reactivation? A laboratory technique in which a sound or smell linked to a memory while awake is replayed during sleep to reactivate that memory. It has been used to strengthen learning and, in the 2026 study, to steer dreams. [2] [1] It needs brain monitoring to time the cues, so it is not yet something to do at home.
Leave the question at the door tonight. Write down whatever is on the sand in the morning, whether or not it looks like a key.
Related reading: How to Remember Your Dreams, Lucid Dreaming: How to Lucid Dream, Step by Step, Your Dreams Aren’t Random, and What Do Dreams Mean?
Lunar Log is a private, local-first dream journal for iOS, Android, and the web. Capture dreams by voice or text before they fade, tag the feeling they left, and see which people, places and patterns keep coming back. It’s free to journal, forever. Get the app.
Sources & further reading
- Konkoly KR, Morris DJ, Hurka K, Martinez AM, Sanders KEG, Paller KA. Creative problem-solving after experimentally provoking dreams of unsolved puzzles during REM sleep. Neuroscience of Consciousness, 2026;2026(1):niaf067.
- Sanders KEG, Osburn S, Paller KA, Beeman M. Targeted memory reactivation during sleep improves next-day problem solving. Psychological Science, 2019;30(11):1616–1624.
- Horowitz AH, Esfahany K, Gálvez TV, Maes P, Stickgold R. Targeted dream incubation at sleep onset increases post-sleep creative performance. Scientific Reports, 2023;13:7319.
- Barrett D. The "committee of sleep": a study of dream incubation for problem solving. Dreaming, 1993;3(2):115–122.
- White GL, Taytroe L. Personal problem-solving using dream incubation: dreaming, relaxation, or waking cognition? Dreaming, 2003;13(4):193–209.
- Wagner U, Gais S, Haider H, Verleger R, Born J. Sleep inspires insight. Nature, 2004;427:352–355.
- Lacaux C, Andrillon T, Bastoul C, Idir Y, Fonteix-Galet A, Arnulf I, Oudiette D. Sleep onset is a creative sweet spot. Science Advances, 2021;7(50):eabj5866.
- Konkoly KR, Appel K, Chabani E, et al. Real-time dialogue between experimenters and dreamers during REM sleep. Current Biology, 2021;31(7):1417–1427.
- Schredl M, Erlacher D. Self-reported effects of dreams on waking-life creativity: an empirical study. The Journal of Psychology, 2007;141(1):35–46.
- Barrett D. Dreams and creative problem-solving. Annals of the New York Academy of Sciences, 2017;1406(1):64–67.
- Aspy DJ. Is dream recall underestimated by retrospective measures and enhanced by keeping a logbook? An empirical investigation. Consciousness and Cognition, 2016;42:181–203.
- Elce V, Bontempi G, Scarpelli S, et al. Individual traits and experiences predict the content of dreams. Communications Psychology, 2026;4(1):69.