Mini-mental state examination (MMSE) template and scoring guide

Key takeaways

The mini-mental state examination is a 30-point screen covering orientation, registration, attention, recall, language, and visuospatial ability.

Scores of 25 to 30 suggest normal cognition, 21 to 24 mild impairment, 10 to 20 moderate impairment, and below 10 severe impairment.

Education, hearing, vision, and first language all move a score, so read the number against the patient in front of you.

One score on its own is weak evidence. Repeat testing against the same patient’s earlier result is what makes the tool useful.

Practice management software like Pabau stores each score in the client record, so the trend sits alongside the rest of the history.

The mini-mental state examination is a brief 30-point screen for cognitive impairment and dementia. Folstein and colleagues published it in 1975, and it is still the most widely used standardized cognitive test in clinical practice.

This guide covers how to administer the test and how to score it. It also covers how to read a result against the patient’s education, hearing, and earlier scores. Download the form above and work through the steps below.

What is the mini-mental state examination?

The mini-mental state examination is a standardized screening tool that scores six areas of cognitive function out of 30. It screens, it does not diagnose. A low score tells you to investigate further, but says nothing about the cause.

The test has held its place for over 45 years. It is short, it needs no equipment, and it has been validated across a wide range of populations. Research from PubMed Central documents that validation work in detail. Most patients finish in 5 to 10 minutes.

It is a scored screen rather than a narrative assessment. Clinicians who also need a written picture of mood, thought, and behavior usually pair it with a mental status exam.

The six cognitive domains it scores

The test covers six domains, each worth a fixed number of points. Knowing what every domain measures helps you administer it accurately and read the result in context.

The six domains add up to 30 points. Record the breakdown rather than the total alone. Two patients scoring 24 can fail entirely different sections, and that difference shapes what you do next.

A patient who loses points only on delayed recall is a different clinical picture from one who loses them on attention. Digital patient records keep each subscore attached to the visit it came from.

How to administer the test: Step by step

Accurate administration comes down to the room, the pacing, and sticking to the scripted wording. Work through these five steps in order.

1. Prepare the room. Run the test somewhere quiet and well lit, with no interruptions. Check that the patient is alert, comfortable, and willing to take part. Allow 5 to 10 minutes.
2. Ask the orientation questions. Ask for the year, season, month, date, and day. Then ask for the country, state, city, building, and floor. One point each, 10 in total.
3. Register three words, then test attention. Name three unrelated objects slowly and clearly, then ask the patient to repeat them straight back. That is worth 3 points. Now ask for serial sevens from 100, stopping after five subtractions, or have them spell “world” backward. Attention carries 5 points.
4. Test recall, language, and copying. Ask for the three words again, scoring 3 points. Then move to naming two objects, repeating a phrase, and following a three-stage command. Finish with reading an instruction, writing a sentence, and copying intersecting pentagons. Those tasks carry 9 points between them.
5. Score as you go. Mark each item as you administer it rather than from memory afterward. Total the six domains, then note anything you saw about effort, speech, or cooperation.

How to score and interpret the result

Scores fall into four severity bands. Age and education shift what those bands mean for an individual patient, so treat the table below as a starting point rather than a verdict.

Education-level adjustment matters here. Lower cut-offs suit patients with limited formal schooling, while higher thresholds apply to university-educated patients. Healthdirect Australia makes the same point about age and baseline function.

Record what you did with the score as well as the number. A screen that raises concern without a confirmed diagnosis is usually coded R41.82, while dementia arising from another diagnosed condition moves to F02.80.

When you refer on, the receiving clinician will usually open with a full diagnostic interview billed under 90791. Sending your domain breakdown with the referral saves them repeating work you have already done.

What can invalidate a score

Before you interpret the number, check that the test measured what you think it measured. Several things depress a score with no cognitive decline behind them at all.

• Hearing loss. A patient who mishears the three words loses points twice, on registration and again on recall. Where hearing is in doubt, a quick Weber test is worth running first.
• Vision and hand function. Reading, writing, and the pentagon copy all need usable sight and a steady hand. Note why those items failed so the total is not read as pure cognitive loss.
• Education and literacy. Limited formal schooling lowers scores on its own, and the reading and writing items assume the patient can read.
• First language. Testing someone in their second language costs points across naming, repetition, and the written tasks.
• Delirium, pain, and sedation. An acutely unwell patient may score low today and far higher next week. Retest once they have settled.

Write these observations on the form itself. A score of 22 with “profoundly deaf, no hearing aid” beside it means something completely different from a bare 22.

Why one score tells you so little

A single result is weak evidence. The number earns its value through comparison, either against a population norm or, far better, against the same patient’s earlier score.

Untreated Alzheimer’s disease typically costs a patient two to four points a year on this test. Measured against that, a two-point drop between annual reviews is ordinary noise. A six-point drop is a reason to act. Without a baseline you cannot tell the two apart.

That is an argument for screening early, while the patient still scores normally. A 28 in a healthy 68-year-old looks like a wasted appointment on the day. Eight years later it can be the most useful number in the chart.

Storing results where you can see them side by side matters as much as running the test. Practices on mental health EMR software can chart every score against its visit date. Filing the same results as scanned paper turns each comparison into a manual hunt.

Whoever sees the patient next also needs to know what was decided. A case management note captures the referral, the family conversation, and the review date in one place.

Where the test falls short

The test is quick and well validated, but it has known blind spots. Understanding them keeps you from leaning on it harder than it can bear.

• Ceiling effect. Highly educated patients can score 30 while already declining, so early impairment slips through unnoticed.
• Barely tests executive function. Planning, abstraction, and set-shifting go almost unmeasured, and they are often the first abilities to weaken.
• Weak for mild cognitive impairment. Sensitivity drops sharply at the early end of the spectrum, which is exactly where MoCA performs better.
• Copyright. PAR Inc. owns the instrument, so check your licensing position before reproducing it at scale.

Score bands are also not disease stages. Where a diagnosis is already in place, a dementia staging chart describes day-to-day function far better than a number does.

MMSE vs MoCA: Which one to reach for

The Montreal Cognitive Assessment (MoCA) is the usual alternative. Both take about the same time and both score out of 30, but they are tuned for different questions.

For routine primary care screening the mini-mental state examination is still the default. It is quick, it needs nothing but a pen, and every clinician who reads the note already knows the scale.

Where the question is early decline, the Montreal cognitive assessment catches more. Psychology practices often run it as standard for that reason.

Who can administer it?

No specialist qualification is required. Primary care physicians, practice nurses, mental health nurses, clinical psychologists, occupational therapists, and allied health professionals can all run the test competently after basic training.

What matters is correct administration and sound interpretation, not job title. Some practices train non-clinical staff to administer it under supervision, which works as long as a clinician reads the result.

In an older patient, cognition rarely travels alone. Many practices pair the screen with a functional reach test for balance, so falls risk and cognitive risk are captured at the same visit.

How Pabau keeps cognitive scores in one place

Running the test on paper and typing the total into the record afterward costs a few minutes per patient. The domain breakdown usually never makes the journey, and neither do the notes on hearing or fatigue.

Practice management software like Pabau lets you build the form once and complete it in the client record during the appointment. Digital forms hold every item, its score, and your observations together.

Once the result sits in the record it becomes searchable. You can pull every patient screened in the past year and set a recall for the next review. Putting this year’s score beside the one from two years ago takes seconds.

That is what turns a snapshot into a trend line, so you can act on a six-point slide instead of noticing it three appointments late.

Keep every cognitive score in the client record

Record MMSE results in Pabau as you run the test, compare them against earlier scores, and set the next review date. No retyping, no scanned paper to hunt through.

Book a demo

Conclusion

The mini-mental state examination earns its place because it is fast, needs no equipment, and is understood by everyone who will read the note. It will not catch subtle decline, and it will penalize a patient for poor hearing or thin schooling if you let it.

So run it, but record it properly. Note the domain breakdown, note anything that could have depressed the score, and set the date for the next one.

A first score is a baseline, not a finding. The second one is where the clinical value shows up. Book a demo to see how Pabau keeps cognitive scores, forms, and review dates in one client record.

Continue your research

Need a full psychiatric assessment structure? Psychiatric evaluation template gives you the interview framework and documentation layout for a complete assessment.

Supporting a patient after diagnosis? Dementia worksheets offer structured activities and orientation exercises you can send home with the family.

Screening older patients for falls risk? Elderly mobility scale scores balance and transfers in the same short appointment as a cognitive screen.

Ruling out low mood behind a low score? Depression self-assessment helps separate depressive symptoms from cognitive decline before you refer.

Building a mental health review from scratch? Psychiatric review of systems covers the symptom checklist that belongs alongside a cognitive screen.

Frequently asked questions