ACE-III: Addenbrooke's Cognitive Examination III

What this assessment is

The Addenbrooke's Cognitive Examination III, or ACE-III, is a clinician-administered cognitive screening examination.

It uses 21 tasks to sample attention, memory, verbal fluency, language, and visuospatial ability. The five domain scores add to a total from 0 to 100.

Higher scores indicate better performance on the tasks.

ACE-III can identify a pattern that needs further assessment. It cannot diagnose dementia, determine its cause, or replace a full clinical evaluation.

This page covers ACE-III. It does not cover the earlier ACE-R or the shorter Mini-ACE.

Who it was designed for

ACE-III was developed for adults being assessed for possible cognitive impairment or dementia.

The original validation study involved people with Alzheimer's disease, frontotemporal dementia, other forms of cognitive impairment, and healthy controls. Later studies have used it in other clinical and community groups.

That history matters. Performance can change with age, education, language, culture, sensory ability, and the reason for referral.

Use a validated language version that fits the person. Do not translate a task during the examination or substitute locally familiar words without evidence that the modified form remains valid.

The official FRONTIER ACE-III resource provides current forms, translations, training information, and reference data.

How it is administered

A trained clinician gives the examination in a quiet setting. Completion usually takes about 15 to 20 minutes.

The person answers spoken questions and completes verbal, written, drawing, memory, and visual tasks. The examiner follows the standard wording and scoring criteria.

Training is important. Small changes to timing, prompts, repetition, or acceptable responses can change the score.

Before interpreting the result, record factors that may have affected performance. These include:

• hearing or vision difficulty
• limited proficiency in the test language
• speech, reading, writing, or movement limitations
• pain, fatigue, anxiety, or low mood
• delirium, acute illness, sleep loss, or medicine effects

Context matters.

Use the complete form. No verified general rule supports estimating an ACE-III total from missing tasks.

Do not guess.

What it measures

ACE-III reports five domain scores:

• attention, up to 18 points
• memory, up to 26 points
• verbal fluency, up to 14 points
• language, up to 26 points
• visuospatial ability, up to 16 points

The domains give more detail than the overall total alone. Two people can earn the same total through different patterns of strengths and difficulty.

Domains matter.

The examination samples cognitive performance at one time. It does not directly measure whether a person can manage medication, finances, transport, meals, or personal care.

The MMSE is a shorter 30-point cognitive screen. MoCA 8.1 and SLUMS also use 30-point totals but contain different tasks and scoring rules.

These scores are not interchangeable.

How scoring works

The examiner scores each task with the official criteria. Points add within each domain and then across all five domains.

The maximum total is 100:

• attention: 18
• memory: 26
• verbal fluency: 14
• language: 26
• visuospatial ability: 16

Higher totals indicate better performance on the sampled tasks.

Report the overall score and all five domain scores. A total without the domain pattern loses useful information.

Keep both.

Do not convert the score to a percentage of normal cognition. A result of 80 does not mean that a person has retained 80 percent of cognitive ability.

Do not replace a missing task with zero, calculate a percentage from answered tasks, or use a scoring rule from ACE-R or Mini-ACE.

How to interpret an ACE-III score

Some studies have examined thresholds near 82 and 88. These values are screening references, not universal diagnostic boundaries.

Scores are not diagnoses.

A Cochrane review found that much of the threshold evidence came from specialist case-control studies with a high risk of bias. The review did not support using ACE-III alone to confirm or exclude dementia.

Interpret the total against suitable reference data for the person's language, education, age, and setting. Then inspect the five domains.

A lower score can support further investigation when it matches reported change or functional difficulty. A higher score does not rule out early or domain-specific impairment.

When tracking change, use the same version, language, setting, and administration method. Practice effects and changes in health can affect repeat results.

The useful question is not only whether the number crossed a cutoff. Ask what changed, which domains contributed, and whether the pattern fits the wider history.

What the score cannot tell you

ACE-III cannot identify the cause of low performance.

It cannot distinguish dementia from delirium, depression, medicine effects, sleep loss, neurological illness, limited education, or a language mismatch by itself.

The score cannot establish decision-making capacity. Capacity is specific to a decision and requires a separate assessment.

It also cannot measure every part of cognition. Brief tasks may miss subtle difficulty, especially when a person has strong education or compensatory strategies.

Do not use the result as a standalone reason to start or stop treatment, restrict independence, or give a diagnosis.

Combine it with history, informant evidence, daily function, physical and neurological assessment, medicines, laboratory tests, imaging when indicated, and follow-up.

Evidence and limitations

ACE-III was designed to retain the broad domain coverage of earlier Addenbrooke's examinations while replacing content that depended on the MMSE.

The original study reported good agreement with ACE-R and useful discrimination in its clinical samples. That does not guarantee the same accuracy in primary care, community screening, different languages, or populations with lower dementia prevalence.

Education and culture affect several tasks. Translation requires more than changing words. Stimulus familiarity, language structure, and local norms can alter performance.

The total can also hide a focal problem. Always retain the domain scores and the circumstances of testing.

Use ACE-III as one structured part of an assessment. A diagnosis still depends on evidence of cognitive change, its effect on daily life, and the clinical explanation for that change.