How to Use a Memory Palace, Step by Step
In one study, beginners went from recalling about 25 words of a 72-word list to about 62 after six weeks of daily practice.
How a memory palace works
A memory palace, also called a mind palace or the method of loci, uses a route you know to hold a list. You place a vivid image of each thing to remember at a stop along it. Then you walk the route in your head to get them back in order.
A review of 13 trials found it improved recall by a medium amount: a real benefit, though not a huge one. Most of those trials had design weaknesses, which makes the size of the benefit less certain.
It takes practice, and it suits lists and ordered facts more than understanding. Below, you'll build your first palace and try it on a shopping list.
Build your first memory palace
I first came across the memory palace in Joshua Foer's book Moonwalking with Einstein. I dived into learning it and used it to smash all my university exams.
There are five steps. This section covers the first two: picking a route and fixing your stops.
The memory palace in five steps
- Choose a routePick a path you know well, such as your walk to work
- Fix the stops in orderPick about 10 spots and always walk them the same way
- Make an image at each stopPicture each item doing something with its spot
- Walk it to recallGo through the route in your head and name what you find
- Come back to it laterRecall it again a few days on, without looking first
Step 1: Pick a route you can picture
Think of a walk you do often, such as from your front door to work, the shops or the station. You need to be able to picture it with your eyes closed, from start to finish, without having to think hard.
A route you travel may work better than a walk round your house. In two experiments, people who set their stops along their route to work recalled more than people who used a route around their home. The study didn't say by how much.
And it doesn't have to be a place you've known for years. In one study, beginners who used a virtual building they'd explored only briefly on a computer did as well as people using a very familiar place.
Step 2: Choose 10 stops, in order
Now walk the route in your head and pick 10 spots along it that you'd notice anyway: your front door, a gate, a bus stop, a shop window. Choose spots that look different from each other, so they don't blur together.
Then fix the order. Walk through the 10 stops from start to finish a couple of times, saying them out loud or writing them down, until you can list them without hesitating. Always walk them the same way round. Your palace is now ready to use.
Why 10? No study we found sets a number, and the trials used lists of anything from 11 to 72 words. Ten is easy to hold in your head and enough for a shopping list. Once it feels easy, add more stops.
How to put things in your memory palace
Once the route is set, the image you leave at each stop does the work. The evidence most supports one rule: make the item do something with the spot.
In one older experiment, pictures of two things interacting helped people recall them better than no pictures. Pictures of the same two things not interacting made recall worse than none. In a memory palace, the two things are the item and its stop.
The same item at the same stop, pictured two ways
- The picture
- Not interacting: Eggs sit by the front doorInteracting: Eggs crack against the door
- Recall
- Not interacting: Worse than no pictureInteracting: Better than no picture
Most guides tell you to make every image as weird as possible: the more flamingos in tutus, the better. The evidence is less keen. In one set of experiments, bizarre images helped only when they were mixed with ordinary ones in the same list. Whole lists of bizarre images did no better than ordinary ones.
So weirdness seems to work like a fire alarm: brilliant once, useless if it's going off all day. Save the odd image for where it counts, and let interaction do the heavy lifting.
Then make each scene as rich as you can, with more than just sight. Hear the eggs crack against the door, feel the cold milk pouring over the gate, smell the coffee spilling out of the office door. Sensory detail gives a scene more to stand out with, and standing out is what seems to help.
That's our suggestion rather than a tested rule. In one set of experiments, how vividly people pictured things, and how good they were at it, didn't explain why interacting images helped. People with little or no mental imagery gained as much as everyone else. So add the detail, but get the interaction right first.
One more thing I'd add from my own use: turn the walk into a story. As you move from one stop to the next, let each scene lead into the next one, so the eggs on the door splash onto the doormat where the loaf is wiping its feet.
The route keeps the order, and the story gives you a second thread to follow if a stop goes blank.
A worked example: a shopping list on your walk to work
Here's a made-up example to try in your head. You need 10 things from the shop, and your walk to work passes 10 spots you know.
A made-up example: ten items, ten stops
- EggsFront door
- BreadDoormat
- MilkFront gate
- BananasBus stop
- CheesePost box
- TomatoesCrossing
- RiceCafé
- SoapPark bench
- OnionsStation
- CoffeeOffice door
| Stop | The scene |
|---|---|
| 1. Front door | Eggs crack against the door as you open it |
| 2. Doormat | A loaf wipes its feet on the mat |
| 3. Front gate | Milk pours over the gate like a waterfall |
| 4. Bus stop | A bunch of bananas waits for the bus |
| 5. Post box | A block of cheese is being posted, and it's stuck |
| 6. Crossing | Tomatoes roll across when the light turns green |
| 7. Café | The barista pours rice into a cup |
| 8. Park bench | The bench is covered in foam, and someone is scrubbing it |
| 9. Station | An onion cries at the ticket barrier |
| 10. Office door | Coffee beans pour out as the door opens |
Now look away and walk the route. At each stop, ask what's happening there. The route tells you where to look next, so you only have to remember what's at each spot, not what comes after it.
How to recall it and make it stick
To get the images back, walk the route forward in your head, name what you find at each stop, and only then check your list.
Then come back to it. A large review of study techniques rated testing yourself and spacing out practice as the two most useful study habits. It didn't test the memory palace, but the two fit together naturally: recall the route without looking, check, and walk it again a few days later.
Can you reuse a memory palace?
Yes, though how far you can push it is less clear. In one lab study, using the same stops for three lists in a row didn't cause more mix-ups than other methods did.
What did cause mix-ups was content: lists of things from the same category interfered with each other whatever method people used.
That study covered only three lists, and a common report on memory forums disagrees with it. People describe old images lingering at a stop and getting in the way of new ones, which they call "ghosting" (one thread, another, a third). A haunting problem is, admittedly, very on-brand for a palace.
Their usual fix is to keep separate palaces for things they want to keep, and rotate several for short-term practice. These are anecdotes rather than evidence.
Still, the six-week trial took a similar line: people built four routes and rotated between them to limit carry-over. So keep a few routes on the go, and give lists on related topics different routes.
How to use a memory palace for exams
The method stores things to recall. It doesn't help you understand them, so learn a topic first. Then use a palace for the parts you must recall in order: the steps of a process, the points of an essay plan or a list of terms.
That's how I used it at university. I turned every fact I possibly could into an image and stored them on a route around the campus. In my final-year exams, I got the highest mark in my year twice.
A campus route works for other people too. In a class demonstration, students who built a palace on campus locations recalled more of a grocery list afterwards, though there was no comparison group.
Turning abstract material into images
Exam material is often abstract, so turn each item into something you can picture. Use what it sounds like, what it looks like or something it reminds you of. Here's an example with made-up scenes: the eight planets in order from the Sun, on a second route around your home.
| Stop | Planet | The scene |
|---|---|---|
| 1. Kitchen door | Mercury | A thermometer is jammed in the cat flap, boiling over |
| 2. Hallway mirror | Venus | A statue of Venus is admiring herself in the mirror |
| 3. Stairs | Earth | A globe bounces down the stairs |
| 4. Landing | Mars | A red toy soldier marches across the carpet |
| 5. Bathroom | Jupiter | A giant striped beach ball is jammed in the bath |
| 6. Bedroom | Saturn | The bed has a ring spinning round it like a hula hoop |
| 7. Wardrobe | Uranus | A blue ball rolls out on its side from between the coats |
| 8. Window | Neptune | A trident smashes through the glass, and seawater pours in |
It starts at the kitchen door, not the front door, so it doesn't share stops with the shopping list.
Reviewing before the exam
Use one palace per topic or paper, so related lists don't sit on the same stops. Then test yourself the way we described above, spread over several days rather than the night before.
You don't have to schedule that by hand. Anki is a flashcard program that times each review for when you're likely to forget, free on computers and Android. Memrise is a language app with a "smart review" feature, and you can start for free.
For vocabulary, the study-techniques review rated the keyword trick as low utility: linking a new word to a familiar word that sounds like it. No study we found tested the palace for learning a language, so use a palace there only for a fixed list of words, with an app doing the testing.
What the classroom studies show
They're encouraging but limited. Medical students taught a hormones topic with lectures plus a session building memory palaces scored better on assessments than students given lectures plus self-study. But the palace group also had an extra teacher-led session, so the palace can't take all the credit.
How you take material in matters too. In one study, the method helped students remember passages better when they heard them read aloud than when they read the text themselves. So if you can, read your notes aloud while you place them.
How long it takes to get good
Some benefit can come early, as in the campus class demonstration above. The big gains, though, came the unglamorous way. In a six-week training study, adults new to the method practised at home for 30 minutes a day, on 40 days. That's about 20 hours by our sum. Not exactly a life hack.
Their recall of a 72-word list, tested about 20 minutes after studying, went from about 25 words to about 62. Four months later they took the test again and still got about 50.
| Memory palace, before | |
|---|---|
| Memory palace, after, highlighted | |
| Dual n-back, before | |
| Dual n-back, after | |
| No training, before | |
| No training, after |
A group that played a brain-training game for the same time went from about 31 to 42. A group that didn't train went from about 29 to 36.
The groups started at slightly different levels and had only 16 or 17 people each, so treat these as rough. And the test used word lists, the same kind of material people practised on.
Being good at it doesn't take a special brain. A study of memory competitors found they weren't unusually clever and had no unusual brain structure. What set them apart was that they used a spatial strategy, the same one you've just learned.
Will it improve your memory overall?
Mostly, it improves what you practise: lists. The game in that chart was dual n-back, where you track squares on a grid and spoken letters, and say when one matches the one a few steps back. It trains holding things in mind and updating them.
A palace trains a way of storing a list, which likely explains why it won so easily on a list test.
A 2021 study of 148 young adults compared the two head to head. Both groups got better at the task they'd trained on and at repeating back strings of digits. Only dual n-back also improved a second untrained test: spotting what had changed between two quick displays. That kind of carry-over is what researchers look for.
In my experience, it worked the same way. I'd struggled with my memory since I was a kid, and I used the memory palace a lot to cope. It was helpful and cool to use, but it didn't fix the underlying problem.
There's a fair case against that view, though. A strategy that reliably works may be all most people need, and the trial's gains were still there four months on.
Even so, I'd tell a friend it's worth diving into and trying for yourself. If nothing else, it makes you very aware of how your own memory works: how much your senses, emotions and stories shape what sticks.
Neurobuff's Portals is our version of dual n-back. The studies here used their own training software, not ours. You can play one free session of the daily featured game, or every game anytime with Premium.
Our guide on whether you can improve working memory covers holding things in mind. Our overview of whether brain training works sets out how far gains carry over to daily life.
Who a memory palace doesn't suit
If you mainly need to understand material rather than recall it, your time is better spent elsewhere. The same goes if you won't put in a few weeks of practice. For something you'll only need once, a note or your phone is easier.
If routes and buildings don't click for you, there are other scaffolds to hang a list on. In one study comparing them, body parts worked as well as a familiar route for remembering lists in order, and personal stories also beat having no scaffold.
How vividly people said they could picture things didn't predict who did well. So if you don't see images clearly in your head, you aren't ruled out.
The studies disagree on routines. In that study, routine activities, like the steps of an everyday task, worked less well than a route or the body.
But in another, personal memories and the steps of making a sandwich did about as well as a palace set in people's homes. People did seem to find some methods easier to master than others.
Try a body list next time you need ten things in order. Use the top of your head, forehead, eyes, nose, mouth, neck, shoulders, chest, stomach and knees, with one item at each. If it sticks better than your walk to work, use that.
Bouffard, N., Stokes, J., Kramer, H. J., & Ekstrom, A. D. (2018). Temporal encoding strategies result in boosts to final free recall performance comparable to spatial ones. Memory & Cognition, 46(1), 17–31. doi.org/10.3758/s13421-017-0742-z
Cornoldi, C., & De Beni, R. (1991). Memory for discourse: Loci mnemonics and the oral presentation effect. Applied Cognitive Psychology, 5(6), 511–518. doi.org/10.1002/acp.2350050606
Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4–58. doi.org/10.1177/1529100612453266
Kluger, F. E., Oladimeji, D. M., Tan, Y., Brown, N. R., & Caplan, J. B. (2022). Mnemonic scaffolds vary in effectiveness for serial recall. Memory, 30(7), 869–894. doi.org/10.1080/09658211.2022.2052322
Legge, E. L. G., Madan, C. R., Ng, E. T., & Caplan, J. B. (2012). Building a memory palace in minutes: Equivalent memory performance using virtual versus conventional environments with the Method of Loci. Acta Psychologica, 141(3), 380–390. doi.org/10.1016/j.actpsy.2012.09.002
Li, W., Zhang, Q., Qiao, H., et al. (2021). Dual n-back working memory training evinces superior transfer effects compared to the method of loci. Scientific Reports, 11, 3072. doi.org/10.1038/s41598-021-82663-w
Maguire, E. A., Valentine, E. R., Wilding, J. M., & Kapur, N. (2003). Routes to remembering: The brains behind superior memory. Nature Neuroscience, 6(1), 90–95. doi.org/10.1038/nn988
Massen, C., & Vaterrodt-Plünnecke, B. (2006). The role of proactive interference in mnemonic techniques. Memory, 14(2), 189–196. doi.org/10.1080/09658210544000042
Massen, C., Vaterrodt-Plünnecke, B., Krings, L., & Hilbig, B. E. (2009). Effects of instruction on learners' ability to generate an effective pathway in the method of loci. Memory, 17(7), 724–731. doi.org/10.1080/09658210903012442
McCabe, J. A. (2015). Location, location, location! Demonstrating the mnemonic benefit of the method of loci. Teaching of Psychology, 42(2), 169–173. doi.org/10.1177/0098628315573143
McDaniel, M. A., & Einstein, G. O. (1986). Bizarre imagery as an effective memory aid: The importance of distinctiveness. Journal of Experimental Psychology: Learning, Memory, and Cognition, 12(1), 54–65. doi.org/10.1037/0278-7393.12.1.54
Qureshi, A., Rizvi, F., Syed, A., Shahid, A., & Manzoor, H. (2014). The method of loci as a mnemonic device to facilitate learning in endocrinology leads to improvement in student performance as measured by assessments. Advances in Physiology Education, 38(2), 140–144. doi.org/10.1152/advan.00092.2013
Thomas, J. J., Ayuno, K. C., Kluger, F. E., & Caplan, J. B. (2023). The relationship between interactive-imagery instructions and association memory. Memory & Cognition, 51(2), 371–390. doi.org/10.3758/s13421-022-01347-6
Twomey, C., & Kroneisen, M. (2021). The effectiveness of the loci method as a mnemonic device: Meta-analysis. Quarterly Journal of Experimental Psychology, 74(8), 1317–1326. doi.org/10.1177/1747021821993457
Wagner, I. C., Konrad, B. N., Schuster, P., Weisig, S., Repantis, D., Ohla, K., Kühn, S., Fernández, G., Steiger, A., Lamm, C., Czisch, M., & Dresler, M. (2021). Durable memories and efficient neural coding through mnemonic training using the method of loci. Science Advances, 7(10), eabc7606. doi.org/10.1126/sciadv.abc7606
Wollen, K. A., Weber, A., & Lowry, D. H. (1972). Bizarreness versus interaction of mental images as determinants of learning. Cognitive Psychology, 3(3), 518–523. doi.org/10.1016/0010-0285