The memory palace, also called the method of loci, is the oldest mnemonic technique still in active use. Cicero described it in De Oratore in 55 BCE, attributing it to the poet Simonides of Ceos. Roman senators used it to memorize speeches without notes. Modern memory athletes use it to memorize the order of shuffled decks of cards in under 20 seconds. The 2017 Dresler et al. study in Neuron showed that six weeks of training in the method roughly doubles the average person's recall and produces measurable changes in brain connectivity that mirror those of elite memory champions.

The technique works because it exploits two cognitive strengths the brain happens to be unusually good at. Spatial memory is deeply preserved across evolution; the hippocampus is built around it. Visual imagery creates richer encoding traces than verbal rehearsal, with multiple retrieval cues. By yoking arbitrary information (a shopping list, a sequence of names, a speech) to a familiar spatial walk and a series of vivid images, you convert a difficult memory problem into one the brain handles natively.

This post is a 5-step method anyone can apply, plus the brain-imaging evidence on why it works.

In 30 seconds: Pick a familiar place. Walk a fixed route through it. At each location, place a vivid image that represents what you want to remember. Walk the route mentally to retrieve. Practice the route daily for a week. The technique scales from grocery lists to entire books.

What is a memory palace?

A memory palace is a mental walk through a familiar space, with a fixed sequence of locations (called loci, Latin for "places") at which you mentally place vivid images. The space can be your childhood home, your current apartment, your daily commute, the inside of a familiar grocery store, or any well-rehearsed environment. The route through it must be fixed; same path, same order, every time you walk it.

The information you want to remember is encoded as images that you place at each locus. To memorize a 10-item shopping list, you walk into your kitchen and see (vividly, almost cartoonishly) a giant carton of eggs cracking on the counter. You walk to the dishwasher and find it stuffed with bananas. You move to the fridge and there is a steak nailed to the door. To retrieve the list, you simply walk the route mentally and notice what is at each stop.

This sounds absurd in writing. In practice, it works extremely reliably, because absurdity is exactly what memorable images are made of.

Why does the method actually work?

The neural explanation is well-developed. Maguire et al., 2003, in Nature Neuroscience, scanned the brains of memory champions while they encoded items using the method of loci. The pattern that lit up was striking: heavy activation of the right posterior hippocampus, the medial parietal cortex, and the retrosplenial cortex, the same circuits used for spatial navigation. The champions were not storing arbitrary lists. They were taking spatial walks.

Dresler et al., 2017, took the next step. They trained 23 ordinary adults on the method of loci for six weeks, 30 minutes a day. They compared the trainees to 17 elite memory athletes and to a control group that did no training. Three findings:

  1. Behavioral. Trainees roughly doubled their recall on a 72-word list, from about 26 words pre-training to about 62 words post-training. Memory athletes recalled around 71. The gap between trainees and athletes had largely closed.
  2. Connectivity. Resting-state brain connectivity shifted in trainees toward the pattern observed in memory athletes, particularly between the medial prefrontal cortex and right hippocampus.
  3. Persistence. A four-month follow-up showed that the recall gains persisted, although less than for the elite athletes who continued daily practice.

"Mnemonic training induced widespread reorganisation of brain functional connectivity that increasingly resembled that of expert memory athletes."

Dresler et al., 2017, Neuron

The mechanism is not magic. It is deep encoding. Each item gets bound into a rich multi-modal trace: a spatial cue (the location), a visual cue (the image), a semantic cue (what the image represents), and an effortful generation cue (you constructed the image). Compared with rote rehearsal, which produces a thin, easily-decayed trace, this is structural encoding through and through.

The 5-step method

The method that produces results in the trial literature is straightforward. Follow these steps in order.

Step 1: Choose your palace

Pick a place you know intimately. The single best choice is the home you grew up in or the place you currently live. Your daily commute also works. The criterion is that you can walk it in your mind without effort and that the layout is fixed.

Avoid imagined or unfamiliar locations until you have practiced. Novelty undermines the spatial-memory advantage you are trying to exploit.

Step 2: Define a fixed route

Walk through the space mentally and pick a fixed sequence of distinct locations. Aim for 10 to 20 to start. Each must be visually and spatially distinct (not "on the kitchen counter" three times, but "on the stove," "in the sink," "on the breakfast table"). Walk the route in the same direction every time. Number the loci. Memorize the sequence by walking it five or six times in the first session, before placing anything.

Step 3: Convert what you want to remember into vivid images

This is the step that takes practice. The image at each locus should be:

Step 4: Walk the route to encode

Walk the route in your mind, slowly, in order. At each locus, place the image with deliberate attention. Take 5 to 10 seconds per locus on the first pass. Notice the location, the image, and the interaction.

Then walk it again, faster, in the same order. Then a third time. By the third pass, most loci will be stable. Loci that fail to stick on the third pass need a stronger image; redesign them.

Step 5: Practice retrieval over expanding intervals

Do not study the list again. Test yourself on it. Walk the route mentally an hour later, the next morning, two days later, a week later. This applies the spacing effect and the testing effect simultaneously. Items that survive the week-long retrieval are essentially permanent.

If you forget an item on retrieval, do not look it up immediately. Pause for at least 30 seconds and try to reconstruct from the surrounding loci ("the toaster was on fire, what was on the counter next to it?"). Effortful retrieval, even when partial, strengthens the trace.

What works and what does not

A few things consistently distinguish people who use the method effectively from those who try it once and conclude it does not work.

  1. Reuse the same palace. Beginners worry they will run out of space. They do not. A single 20-locus palace can be reused for thousands of different lists across years; the new images overwrite the old ones, much like working memory does for short-term content. Building dozens of palaces is for advanced users.
  2. Start small. A 10-item shopping list is enough to convince yourself the technique works. A 100-item list is a project for week three.
  3. Spend more time on images than on the route. Most beginner failures are at the imagery step, not the spatial step. The route is the easy part.
  4. Do not try to memorize everything this way. The method works best for ordered, discrete, image-able information. For abstract material (philosophical arguments, statistical formulas), retrieval practice and spaced repetition are stronger tools.
  5. Practice daily for at least two weeks. The Dresler trial dose was 30 minutes a day for 40 days. Less time produces real but smaller gains. The technique gets dramatically better with practice.

What this is good for, and what it is not

The method of loci is excellent for:

It is less useful for:

For the broader picture of how memory actually works (and why the method exploits it so effectively), see our memory 101 guide.

The memory palace is one of the few cognitive techniques where the historical claims hold up almost exactly as advertised. The Romans were not exaggerating.