The Method of Loci is the oldest memory technique in continuous use anywhere in the world. The Greek poet Simonides of Ceos is credited with discovering it around 477 BCE — the story, told by Cicero in De Oratore, goes that Simonides identified the bodies of guests at a collapsed banquet by recalling where each had been seated. From that observation came the idea that spatial memory could be deliberately recruited to carry other kinds of information. Two and a half thousand years later, every memory champion at the World Memory Championships uses some variant of the same technique, and it has held up under modern scanner-based investigation in ways most “brain training” claims have not. This article explains why it works, what the research shows, and how to build a usable memory palace in an evening.

What the technique actually is

A memory palace — locus, plural loci, meaning “place” — is a familiar physical route, mentally walked, with items to be remembered placed vividly at specific stations along it. The standard demonstration goes like this. Imagine your front door. You’re trying to remember a shopping list: eggs, bread, milk, coffee, apples. At the front door, picture a giant egg cracking and spilling yolk down the porch. Walk into the entryway: a loaf of bread is wedged into the umbrella stand, somehow too large to fit. In the kitchen, milk is pouring out of the ceiling. By the stairs, a coffee bean the size of a basketball is rolling down toward you. In the living room, the apples are stacked precariously on the sofa, about to fall. You don’t need to “remember” the list. You walk the route and the items are simply there, where you put them.

The reason this works has been argued about for centuries and finally has a half-decent neural explanation. Spatial navigation is one of the oldest memory systems in the mammalian brain — it predates language by roughly the entire history of vertebrates — and humans have an enormous amount of dedicated cortical and subcortical machinery for it. The hippocampus and the entorhinal cortex contain specialised “place cells” and “grid cells” that fire when an animal occupies particular locations in space, a finding that won John O’Keefe and the Mosers the 2014 Nobel Prize in Physiology or Medicine. The Method of Loci hijacks this system. By tagging information to vivid spatial locations along a familiar route, you piggyback abstract material onto the most heavily reinforced memory infrastructure you have.

What the brain scans show

Most claims about memory techniques don’t survive contact with fMRI. The Method of Loci is unusual: it does.

In a 2003 study published in Nature Neuroscience, Eleanor Maguire and colleagues at University College London scanned ten of the top finishers from the World Memory Championships while they performed a memory task. The memory athletes did not have anatomically larger brains than the matched control group, and they did not score higher on general intelligence measures. What differed was which regions they recruited. When memorising lists of digits, faces, or snowflakes, the memory athletes lit up brain regions associated with spatial navigation and route-learning — including the right posterior hippocampus and retrosplenial cortex — to a far greater degree than controls performing the same task. Nine of the ten athletes reported using the Method of Loci. The brain scans matched what the athletes said they were doing.

A 2017 study by Martin Dresing and colleagues, published in Neuron, took the next step. They trained ordinary undergraduates in the Method of Loci for forty days and measured their performance and connectivity before and after. After training, the students’ performance on word-list recall improved roughly fourfold, and the connectivity pattern in their resting-state fMRI began to resemble that of trained memory athletes — particularly in connections between the medial prefrontal cortex and the hippocampal system. Crucially, the gains were durable: a follow-up four months later showed the recall advantage was still largely intact, even for students who had stopped practising.

Two findings worth holding on to from this body of work. First, the Method of Loci doesn’t require anything unusual about your brain — ordinary people learn it and gain substantial benefit. Second, the gains persist; this isn’t a state-dependent trick that evaporates when you stop using it.

Why it works (the working theory)

Cognitive psychologists offer several overlapping accounts of why placing information along a spatial route produces such reliable recall. The most economical synthesis runs roughly as follows.

You’re using a stronger memory system to scaffold a weaker one. Verbal short-term memory is famously limited — George Miller’s 1956 paper “The Magical Number Seven, Plus or Minus Two” pinned the working-memory span at roughly seven items, give or take, and subsequent work by Nelson Cowan in 2001 revised that downward to closer to four. Spatial memory, by contrast, can carry orders of magnitude more information without collapsing. If you’ve ever navigated a familiar city by memory, considered the layout of your childhood home, or recalled the route to a friend’s house, you’ve already demonstrated a spatial memory capacity that dwarfs your verbal one. The Method of Loci converts the verbal problem into a spatial one.

You’re forcing elaboration. Memory researchers since Craik and Lockhart (1972) have shown that information processed at a deeper, more elaborated level — connected to existing knowledge, encoded with sensory detail, given meaning — is recalled far better than information processed shallowly. Constructing a vivid image of a cracking egg on your porch is, by definition, deep elaborative encoding. You’re not just hearing the word “egg”; you’re visualising it, placing it spatially, often adding kinetic and auditory detail, all of which means the encoded representation has many more retrieval pathways back to it.

You’re providing distinctive retrieval cues. Endel Tulving’s encoding-specificity principle — formulated in the 1970s and still load-bearing in the modern field — holds that recall is best when the cues present at retrieval match the cues that were active at encoding. Each station along your memory palace acts as a built-in retrieval cue. You don’t have to think “what’s next on the list”; you think “what’s at the umbrella stand,” and the bread is simply there.

What the technique is good for — and what it isn’t

Memory palaces excel at a specific class of problem: remembering a sequence of distinct items where the sequence and the items both matter. Shopping lists, the order of US presidents, the planets, the periodic table by atomic number, a speech’s main points, the items in a magic trick, a phone number, an exam-day list of formulas. They are good at things you’d otherwise need to write down.

They are not magic, and several common applications work poorly.

Conceptual understanding doesn’t transfer. Memorising the order of operations in a maths problem does not give you mathematical reasoning. Memorising the names of the Renaissance painters does not give you a feel for what made the period distinctive. Recall of facts is necessary but not sufficient for understanding; the Method of Loci helps with the recall component but not the comprehension one.

Material without internal structure is harder. Memorising a poem stanza-by-stanza fits the technique cleanly; memorising a free-floating set of disconnected words actually requires more invented structure than the technique naturally provides. (Memory athletes spend significant training time pre-assigning images to specific cards or digits — building, in effect, an enormous personal dictionary.)

Long-term, low-effort retention of large bodies of information still benefits more from spaced retrieval practice. The Method of Loci is excellent for getting a specific set of items locked in; spaced repetition is excellent for keeping a much larger set of items permanently available. Most people who use both treat them as complementary tools — palaces for “I need to know this list, cold, in one week” and spaced retrieval for “I want to have this material at my fingertips for years.”

How to build a working palace in an evening

The temptation when you first read about the technique is to overthink the palace itself. Don’t. The standard advice from competitive memorisers is: pick a route you already know well, walk it mentally a few times to fix the stations, and then start using it. Reactive palaces that take six months to plan never get used; rough-and-ready palaces that you start using on day one tend to stick.

Here is the minimum viable workflow.

  1. Pick a route of about ten to fifteen stations. Your home, walked from the front door to the kitchen via every room you’d ordinarily pass, works for most people. Some prefer a familiar walking route — to a school, a favourite café, a friend’s flat. The criterion is that you can walk it in your head without having to invent details.

  2. Number the stations. Front door = 1, hallway = 2, kitchen sink = 3, fridge = 4, and so on. Stick with the same order every time. The numbering is what lets you say “what was at station seven” rather than having to walk the whole route to find a specific item.

  3. Place the first item as vividly and absurdly as you can. Memory athletes converge on a similar principle: the more absurd, kinetic, multi-sensory, and emotionally charged the image, the better it sticks. A loaf of bread sitting tidily on the counter is fragile. A loaf of bread that has crashed through the ceiling and is leaking flour onto a screaming cat is durable. Lean into the cartoon physics.

  4. Walk the route, mentally, three times right after placing the items. Don’t skip this step. The first two walk-throughs will feel laborious; by the third, the items will start “appearing” without effort.

  5. Test yourself an hour later, then the next morning, then a week later. This is just the testing effect (Roediger and Karpicke, 2006) applied to your palace contents. Spaced retrieval, even just three or four times, dramatically strengthens what the palace already holds.

If you’re using the palace for short-term content — a shopping list, a meeting agenda — you can reuse the same palace next week with new items at the same stations. The old contents fade quickly when nothing reinforces them. For long-term content, dedicate a palace and don’t reuse it.

Where the technique fits in a broader practice

The Method of Loci is one technique among several worth using, not a complete approach to learning. Spaced retrieval practice (Cepeda et al., 2008) is the most heavily replicated effect in the cognitive-psychology of learning and probably the single highest-return habit you can adopt. The testing effect (Roediger and Karpicke, 2006) — actively recalling material instead of rereading it — adds another large boost. Memory palaces sit alongside these as a specialised tool for sequenced, item-by-item recall.

A useful mental model: spaced retrieval is the gym. The testing effect is good form. The Method of Loci is a specialised piece of equipment you reach for when the lift requires it. None of them replaces the others; together, they’re a coherent practice.

If you want to put any of this to work tonight, our eight evidence-based methods for improving general knowledge sets out a routine you can run weekly, and our study techniques guide goes deeper on the retrieval and spacing side. For a quick recall check, pick a category and play a round — every active recall counts.

Sources and further reading

  • 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.
  • Dresler, M., Shirer, W. R., Konrad, B. N., et al. (2017). Mnemonic training reshapes brain networks to support superior memory. Neuron, 93(5), 1227–1235.
  • O’Keefe, J., & Dostrovsky, J. (1971). The hippocampus as a spatial map. Brain Research, 34(1), 171–175.
  • Miller, G. A. (1956). The magical number seven, plus or minus two. Psychological Review, 63(2), 81–97.
  • Cowan, N. (2001). The magical number 4 in short-term memory. Behavioral and Brain Sciences, 24(1), 87–114.
  • Craik, F. I. M., & Lockhart, R. S. (1972). Levels of processing: a framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11(6), 671–684.
  • Tulving, E., & Thomson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80(5), 352–373.
  • Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255.
  • Cepeda, N. J., Vul, E., Rohrer, D., Wixted, J. T., & Pashler, H. (2008). Spacing effects in learning: a temporal ridgeline of optimal retention. Psychological Science, 19(11), 1095–1102.