What Is the 3-Word Test for Dementia?

The “3-word test” refers to a short memory task in which a person is asked to remember three unrelated words, perform a brief distraction activity, and then recall those words a few minutes later. It is most commonly encountered as part of the Mini-Cog, a rapid dementia screening tool used in clinics worldwide. The test takes only a few minutes and requires no special equipment, which is why it has become one of the most widely used quick checks for cognitive problems. But the simplicity of the task can be misleading, both about what a good or bad score actually means and about how much the test can tell you on its own.

How the Test Works Step by Step

In the standard Mini-Cog version, a clinician reads three unrelated words aloud, such as “apple, watch, flag,” and asks the person to repeat them immediately. This initial repetition confirms the person heard and registered the words. Then comes a deliberate interruption: the person is asked to draw a clock face showing a specific time. The clock-drawing task typically takes a minute or two, which creates a short delay. After the clock is finished, the person is asked to recall the three words from earlier. One point is awarded for each word recalled correctly, for a maximum of three. The clock drawing itself is scored separately: a correctly drawn clock earns two points, and any significant error earns zero. The combined score ranges from zero to five, and a total of two or below is considered a positive screen for possible dementia.

The whole thing takes about three minutes. That speed is the point. Longer cognitive batteries can take 30 to 90 minutes and require trained neuropsychologists to administer. The Mini-Cog was designed so that a primary care physician, a nurse, or even a medical assistant could run it during a routine office visit without disrupting the flow of the appointment. The tradeoff is precision: a brief screen gives you a quick signal, not a diagnosis.

Why Forgetting Three Words Matters

Remembering a short list of words after a delay sounds trivial, but it taps into a specific brain function that Alzheimer’s disease attacks early: the ability to form and retrieve new memories. The hippocampus, a seahorse-shaped structure deep in the temporal lobe, is critical for converting short-term impressions into stable, retrievable memories. In Alzheimer’s, the hippocampus is among the first regions to accumulate damage.

Research from the Nun Study, a long-running investigation of aging and brain health in a group of Catholic nuns, found that the volume of the hippocampus was strongly linked to how many words a person could recall after a delay. This association held regardless of age and education, and it persisted even when researchers accounted for the amount of Alzheimer’s-related pathology visible in brain tissue. In other words, hippocampal shrinkage predicted poor delayed recall whether or not the person had already been diagnosed with dementia.

The distinction between encoding and retrieval matters here. Encoding is the process of laying down a memory in the first place; retrieval is fishing it back out later. Research suggests that in normal aging, the bigger problem is encoding: the brain becomes less efficient at deeply processing new information, so the memory trace is weaker from the start. In people who go on to develop Alzheimer’s, encoding failure appears even more pronounced and follows a pattern consistent with prodromal disease, the stage before full-blown dementia becomes obvious.

³⁴

So the three-word recall task is not measuring general forgetfulness. It is probing whether the brain can still take in new information, hold it through a distraction, and pull it back. When that specific ability breaks down, it often signals something beyond the normal wear of aging.

What the Clock Drawing Adds

A three-word recall by itself tests memory. The clock-drawing portion tests something different: executive function, which includes planning, spatial reasoning, and the ability to follow a sequence of steps. Drawing a clock face with the numbers in the right positions and the hands pointing to a specific time requires a surprising amount of coordinated brain work. A person has to recall what a clock looks like, organize the numbers around the circle, and then translate a verbal instruction (“set it to ten past eleven”) into the correct hand positions.

Studies have found that the Mini-Cog, the three-item recall, and the clock-drawing test all perform at acceptable levels in distinguishing people with mild cognitive problems from those with normal cognition, but they measure somewhat different things. The recall targets memory systems, while the clock drawing picks up problems in executive function and visuospatial processing. Combining the two in one short test gives a broader snapshot than either piece alone.

⁵

This is also why the clock-drawing task serves as the delay between hearing the words and recalling them. It is not an arbitrary pause. It forces the person to shift their attention to a completely different task, which means the memory trace of those three words has to survive on its own rather than being kept alive by constant rehearsal. If you simply waited in silence for two minutes, many people would just repeat the words to themselves the entire time, which tells you nothing about actual memory consolidation.

How Accurate Is the Test, Really

The Mini-Cog is designed to be a screening tool, not a diagnostic one. That distinction gets lost in casual conversation. Screening means the test is meant to flag people who should undergo further evaluation. It does not confirm or rule out any particular condition on its own.

The three-word memory portion, taken in isolation without the clock-drawing component, has meaningful limitations. One study that compared a standalone three-word recall task to a more structured memory screen found that the simple three-word version had a sensitivity of about 65% and a specificity of about 85% for detecting Alzheimer’s disease. The more structured test, which added category cues and controlled the learning process, performed substantially better on both measures.

⁶

What those numbers mean in practice is that a plain three-word recall misses roughly a third of people who actually have Alzheimer’s and occasionally flags people who do not. Adding the clock drawing improves the overall performance, but the Mini-Cog is still a blunt instrument compared to a full neuropsychological evaluation. Its value lies in speed and accessibility. In a busy primary care clinic where a 45-minute cognitive battery is not feasible, catching two-thirds of cases in three minutes is still valuable if everyone who screens positive gets referred for proper testing.

When Depression Looks Like Dementia

One of the trickiest problems with any short memory test is that depression can mimic cognitive decline. People who are severely depressed often have trouble concentrating, forget recent conversations, and struggle with tasks that require sustained mental effort. These symptoms can look a lot like early dementia, and a three-word recall test is not great at telling the two apart.

Research on delayed word recall in depressed versus Alzheimer’s patients found substantial overlap between the two groups. Using recommended cutoff scores for detecting Alzheimer’s, about 44% of depressed patients were misclassified as demented based on their free recall scores, and roughly 48% were misclassified based on recognition scores.

⁷

This is a major practical concern. Depression is common in older adults, and so is the overlap between depression and early-stage dementia. A person who screens positive on the Mini-Cog may have dementia, may have depression, or may have both. The test alone cannot sort this out, and treating someone for dementia when they actually have untreated depression, or vice versa, leads to very different outcomes. This is one of the strongest reasons why the three-word test should never be treated as a standalone diagnosis.

Stress and Performance on the Day

Even in people with perfectly healthy brains, stress can temporarily impair memory performance. Research on the effects of acute stress on recall has shown that elevated cortisol, the hormone your body releases under stress, directly interferes with the ability to encode and retrieve information. In one study of military special forces candidates exposed to intense stress, cortisol increases were associated with a significant decline in immediate recall, and the memory effects persisted even 24 hours later during recovery.

⁸

Obviously, a doctor’s office is not a military survival course. But for many older adults, being tested in a clinical setting can be anxiety-provoking. The awareness that you are being evaluated for cognitive decline adds emotional weight to the task. Someone who is nervous, sleep-deprived, or simply having a bad day may perform worse than their true cognitive baseline. Clinicians generally know this and are trained to interpret results in context, but if you or a family member scores poorly, it is worth considering whether situational factors played a role before assuming the worst.

Education and Language Can Shift Results

Short cognitive tests are sensitive to a person’s educational background and language proficiency. Research on a similar word-and-shape recall test found that people with fewer years of schooling performed worse on recall of shapes and on overall scores, even when there was nothing wrong with their cognition.

⁹

The clock-drawing task can also be affected by education. People who are not accustomed to written or drawn tasks may struggle with it for reasons that have nothing to do with dementia. Similarly, when the test is administered in a language that is not the person’s first, word recall becomes harder simply because the words are less familiar or less deeply encoded. Validation studies of translated versions of the Mini-Cog have shown that combining the test with other instruments can improve accuracy, but the results vary across cultures and languages.

¹⁰

These biases do not invalidate the test, but they mean that a score of zero on the three-word recall in a person with limited formal education does not carry the same clinical weight as the same score in someone with a graduate degree. Good clinicians account for this when interpreting results.

The Digital Version May Catch More Cases

One limitation of the traditional Mini-Cog is that the clock drawing is scored in a binary way: the clock is either normal or not. That throws away a lot of information. A clock that is slightly off might reveal subtle problems that a pass/fail grading system ignores. Similarly, the traditional format does not capture how long the person takes, whether they hesitate, or how they approach the task.

A digital version called Digital Clock and Recall (DCR) attempts to fix this by having people draw the clock on a tablet, which records every pen stroke, pause, and correction. In testing, the DCR identified significantly more cases of cognitive impairment than the standard Mini-Cog. Using a Mini-Cog cutoff score of four, the DCR caught 47 additional cases that the standard version missed, while the Mini-Cog missed 87 cases that the DCR flagged. The digital version correctly classified 44 cognitively impaired cases that the traditional Mini-Cog would have let through, compared to only four cases where the reverse was true.

¹¹

The process data, such as how smoothly someone draws, how long they pause between numbers, and whether they self-correct, seems to contain diagnostic information that gets lost when a human scorer simply looks at the finished product and judges it as “correct” or “incorrect.” Digital tools like this are not yet standard in most clinics, but they point to a future where the basic structure of the three-word-plus-clock test stays the same while the scoring becomes much more sensitive.

When People Do Not Realize They Have a Problem

One of the more unsettling features of Alzheimer’s disease is that many people who have it do not recognize their own cognitive decline. This phenomenon, known as anosognosia, is not the same as denial. It is a neurological symptom in which the brain’s self-monitoring systems are impaired, making the person genuinely unaware that anything is wrong.

Research has found that the gap between a patient’s self-assessment and their caregiver’s assessment of their abilities grows wider as Alzheimer’s progresses. In milder stages, patients may be somewhat aware of memory problems, particularly around practical tasks like remembering appointments. But as the disease advances into moderate stages, the discrepancy between what patients believe about their own functioning and what caregivers observe becomes much larger.

¹²

This has direct implications for who ends up taking the three-word test in the first place. The people most in need of screening may be the least likely to seek it out or agree to it, because from their perspective nothing is wrong. Family members who notice a loved one repeating questions, forgetting recent events, or struggling with familiar tasks often face resistance when they suggest a cognitive check. Understanding that this resistance may itself be a symptom, rather than stubbornness, can change how families approach the conversation.

What to Do With the Results

If you or someone you care about takes the Mini-Cog and scores in the concerning range, the single most important thing to know is that this does not mean a dementia diagnosis has been made. A low score means further evaluation is warranted. That typically involves a longer, more detailed cognitive assessment, blood work to rule out reversible causes of memory problems (thyroid issues, vitamin deficiencies, medication side effects), and sometimes brain imaging.

If the score is normal, that is reassuring but not a guarantee. The test’s sensitivity, especially for the word-recall portion alone, means that some people in early stages of decline will pass. If you or a family member have noticed real changes in daily functioning despite a normal screening result, it is entirely reasonable to ask for a more thorough evaluation anyway. Screening tools are designed to work well across large populations. For any individual person, clinical judgment and real-world observations still matter more than a single three-minute test.

The three-word test endures because it captures something fundamental about how memory breaks down in dementia, and it does so quickly enough to be practical. Its value lies not in giving a final answer but in asking the right question at the right time, ideally before significant cognitive ground has already been lost.