What are the recommended cutoff scores for MCI and dementia on the MoCA by age and education?

Checked on December 4, 2025
Disclaimer: Factually can make mistakes. Please verify important information or breaking news. Learn more.

Executive summary

The original MoCA recommends a single cutoff of 26/30 to separate “normal” from cognitive impairment with a one‑point education correction for people with ≤12 years of formal schooling (add 1 point) [1] [2] [3]. Multiple later studies and large datasets show optimal cutoffs vary markedly by age, education and race/ethnicity — reported alternatives include cutoffs near 23–25 for MCI and much lower values for dementia depending on subgroup (examples: ≤23 meta‑analysis, age/education‑specific cutoffs <22 to <28 across strata, and race/ethnicity‑specific MCI cutoffs 23–25 and dementia cutoffs 16–19) [4] [5] [6] [7].

1. The official, simple rule clinicians were taught

The widely cited original MoCA recommendation: a single screening threshold of 26/30 classifies possible cognitive impairment, with an additive correction of +1 point for examinees who have 12 years or fewer of formal education (do not count kindergarten) — this is how the test authors and MoCA materials instruct routine scoring [1] [8] [3] [2].

2. Why many researchers now reject one global cutoff

Multiple validation and normative studies show a fixed 26 cutoff produces many false positives in older and less‑educated people; a meta‑analysis recommended lowering the threshold to 23 to improve overall diagnostic accuracy [4]. Large, racially diverse analyses also report that optimal cutoffs differ by race/ethnicity and education — e.g., optimal MCI cutoffs were 25 in non‑Hispanic Whites, 24 in Hispanics and 23 in non‑Hispanic Blacks; dementia cutoffs ranged 16–19 across groups [6] [7].

3. Age and education stratification yields very different thresholds

Age-and-education stratified work finds the “best” cutoff shifts with both factors. One study that derived cutoffs by age (<55, 55–70, >70) and education (≤12 vs >12 years) recommended maximizing sensitivity/specificity with cutoffs such as <28 for patients <55 with >12 years education, and <22 (≤12 years education) or <25 (>12 years) for those >70 — other groups were best served by <26 [5]. Stroke and older‑adult normative datasets similarly show single cutoffs (21/22 or 25/26) misclassify many healthy older adults, especially those with less education [9].

4. Practical implications: MCI vs dementia thresholds

Across sources the separation between “normal,” MCI and dementia is not universal: the original averages were ~27.4 for normals, ~22.1 for MCI and ~16.2 for Alzheimer’s disease in one study, but later work recommends different cutoffs depending on purpose — e.g., a cutoff ≤24 or ≤22 improved sensitivity/specificity tradeoffs for MCI in some cohorts; meta‑analytic work favored 23 as a compromise [1] [3] [4]. Race/ethnicity analyses place optimal dementia cutoffs as low as 16 for some groups [6].

5. Education adjustment: +1 point is common but imperfect

The simple correction — add 1 point if education ≤12 years — appears in the original and many clinical guidelines [3] [2]. Multiple studies question whether a single +1 adequately compensates for education effects; researchers propose stratified norms, regression adjustments, item‑level modifications or alternative cutoffs to reduce misclassification among lower‑education groups [10] [11].

6. Competing perspectives and tradeoffs

Supporters of a single cutoff argue it is simple and sensitive for screening (and the MoCA creators endorsed 26) [2]. Critics and many empirical studies counter that single cutoffs inflate false positives in older, less‑educated or minority groups; stratified cutoffs improve specificity but can lower sensitivity, so choice depends on whether the clinician prioritizes catching all possible impairment (high sensitivity) or avoiding false alarms (high specificity) [7] [9] [5].

7. How to use this in clinical practice

Available sources recommend using the original cutoff (≤26 with +1 education correction) for broad screening or where high sensitivity is desired, but to interpret results against age‑ and education‑adjusted norms or race/ethnicity–specific cutoffs where available to avoid misclassification [2] [7] [5]. For diagnostic decisions, pair the MoCA with detailed neuropsychological testing rather than relying on a single threshold (not found in current reporting: specific step‑by‑step clinical algorithms beyond these studies; sources emphasize the need for context and confirmatory testing) [3] [12].

Limitations: different cohorts, languages and administration methods produced the reported cutoffs; no single “best” values apply to all settings. The sources above explicitly demonstrate both the original single cutoff [1] [2] and multiple alternative, stratified cutoffs [5] [6] [7] [4]. Use local norms when possible.

Want to dive deeper?
What MoCA score ranges correspond to normal cognition, MCI, and dementia overall?
How should MoCA cutoffs be adjusted for years of education and illiteracy?
Are there age-specific normative MoCA scores for different decades of life?
How do language, culture, and education affect MoCA validity and cutoffs?
What alternative screening tools and cutoffs exist for MCI and dementia besides the MoCA?