This is the widely used 22-item unit-scored version of the MAST. The original 1971 instrument used weighted scoring (range 0-53, cutoff 5 or more); the item content is essentially the same. All questions are lifetime-framed: the MAST screens for a history of alcohol problems and cannot track change.
Yes1
No0
Score
01Do you feel you are a normal drinker? ("normal" - drink as much or less than most other people)
-
02Have you ever awakened the morning after some drinking the night before and found that you could not remember a part of the evening?
-
03Does any near relative or close friend ever worry or complain about your drinking?
-
04Can you stop drinking without difficulty after one or two drinks?
-
05Do you ever feel guilty about your drinking?
-
06Have you ever attended a meeting of Alcoholics Anonymous (AA)?
-
07Have you ever gotten into physical fights when drinking?
-
08Has drinking ever created problems between you and a near relative or close friend?
-
09Has any family member or close friend gone to anyone for help about your drinking?
-
10Have you ever lost friends because of your drinking?
-
11Have you ever gotten into trouble at work because of drinking?
-
12Have you ever lost a job because of drinking?
-
13Have you ever neglected your obligations, your family, or your work for two or more days in a row because you were drinking?
-
14Do you drink before noon fairly often?
-
15Have you ever been told you have liver trouble such as cirrhosis?
-
16After heavy drinking have you ever had delirium tremens (D.T.'s), severe shaking, visual or auditory (hearing) hallucinations?
-
17Have you ever gone to anyone for help about your drinking?
-
18Have you ever been hospitalized because of drinking?
-
19Has your drinking ever resulted in your being hospitalized in a psychiatric ward?
-
20Have you ever gone to any doctor, social worker, clergyman or mental health clinic for help with any emotional problem in which drinking was part of the problem?
-
21Have you been arrested more than once for driving under the influence of alcohol?
-
22Have you ever been arrested, even for a few hours, because of other behaviour while drinking?
-
0-3score item↑↓moveScored locally - nothing leaves this page
0 of 22
0 / 22
MAST0 / 22
This is the widely used 22-item unit-scored version of the MAST. The original 1971 instrument used weighted scoring (range 0-53, cutoff 5 or more); the item content is essentially the same. All questions are lifetime-framed: the MAST screens for a history of alcohol problems and cannot track change.
Scored locally - nothing leaves this page
01Do you feel you are a normal drinker? ("normal" - drink as much or less than most other people)
02Have you ever awakened the morning after some drinking the night before and found that you could not remember a part of the evening?
03Does any near relative or close friend ever worry or complain about your drinking?
04Can you stop drinking without difficulty after one or two drinks?
05Do you ever feel guilty about your drinking?
06Have you ever attended a meeting of Alcoholics Anonymous (AA)?
07Have you ever gotten into physical fights when drinking?
08Has drinking ever created problems between you and a near relative or close friend?
09Has any family member or close friend gone to anyone for help about your drinking?
10Have you ever lost friends because of your drinking?
11Have you ever gotten into trouble at work because of drinking?
12Have you ever lost a job because of drinking?
13Have you ever neglected your obligations, your family, or your work for two or more days in a row because you were drinking?
14Do you drink before noon fairly often?
15Have you ever been told you have liver trouble such as cirrhosis?
16After heavy drinking have you ever had delirium tremens (D.T.'s), severe shaking, visual or auditory (hearing) hallucinations?
17Have you ever gone to anyone for help about your drinking?
18Have you ever been hospitalized because of drinking?
19Has your drinking ever resulted in your being hospitalized in a psychiatric ward?
20Have you ever gone to any doctor, social worker, clergyman or mental health clinic for help with any emotional problem in which drinking was part of the problem?
21Have you been arrested more than once for driving under the influence of alcohol?
22Have you ever been arrested, even for a few hours, because of other behaviour while drinking?
The MAST asks about the consequences of drinking rather than the amount consumed. Its twenty-two yes/no questions cover other people's concern about the respondent's drinking, guilt, blackouts, morning drinking, loss of control, arguments and violence, damage to work, family and friendships, medical complications including liver disease and delirium tremens, arrest for drink-driving or drunkenness, and previous contact with treatment services or Alcoholics Anonymous. Not one item asks how much or how often the person drinks.
That design decision is the key to interpreting the instrument. A high MAST score describes a life in which alcohol has already caused damage, which is close to what older diagnostic frameworks meant by alcoholism and what DSM-5 would place at the moderate-to-severe end of alcohol use disorder. Every question is framed across the lifetime, so a positive result tells you that alcohol has been a problem at some point, not that it is a problem now. The MAST is a case-finding instrument for established alcohol-related harm, not a measure of current consumption and not a monitoring tool.
02 - Origin & purpose
Where it comes from.
Melvin Selzer published the MAST in the American Journal of Psychiatry in 1971, under the title "The Michigan Alcoholism Screening Test: the quest for a new diagnostic instrument". At the time no brief, standardised, self-administered screen for alcoholism existed; diagnosis rested on clinical judgement, and Selzer's aim was a short questionnaire that a non-specialist could give and score. The original instrument had twenty-five items with weighted scoring, each item carrying one, two or five points to a maximum of 53, and a total of five or more was read as probable alcoholism. 1
The weighted scoring proved cumbersome and was widely abandoned in favour of unit scoring, in which each endorsed item counts one point. Several variants followed: a thirteen-item short form (SMAST), a ten-item brief form (bMAST), and a twenty-four-item geriatric version (MAST-G, Blow and colleagues, 1992) with its own ten-item short form, developed because the standard items - work problems, drink-driving arrests - map poorly onto retired lives. 4 The version reproduced on this page is the twenty-two-item unit-scored form distributed by US state health authorities. From the 1990s onward the AUDIT largely displaced the MAST in guidelines, because the AUDIT asks about current consumption and detects hazardous drinking before consequences accumulate; the MAST retains a role where the question is whether alcohol has caused harm. 5
03 - Scoring & cutoffs
How scoring works.
Each of the twenty-two items is answered yes or no, and each endorsement scores one point, giving a total between 0 and 22. Two items are reverse-scored: item 1, which asks whether the respondent feels they are a normal drinker, and item 4, which asks whether friends or relatives think they are a normal drinker. On both, a "no" is the positive response, because denial of normality is the clinically meaningful answer. The calculator handles the reversal, but it matters if you are scoring a paper form by hand.
Interpretation on this version follows three bands: 0–2 suggests no current indication of an alcohol problem, 3–5 is a borderline result warranting further enquiry, and 6 or more indicates probable alcohol dependence and should prompt a full assessment. Readers coming from the original 1971 paper will see a cutoff of 5, but that figure belongs to the weighted 0–53 scoring, where five points could be a single item; the two thresholds are not interchangeable. Because every item is lifetime-framed, a total cannot be compared across visits and a "recovered" patient with a long history will continue to score highly. Score the instrument as a history, and always ask separately about current drinking.
Score
Severity
Interpretation
0–2
No apparent problem
Below the screening threshold on this version.
3–5
Early or middle problem drinker
Suggestive of an emerging alcohol problem; explore drinking history clinically.
6–22
Problem drinker
Consistent with alcohol dependence on this screen; full clinical assessment indicated.
04 - Validation evidence
How well it performs.
Selzer's original report described a screen that separated a group with an established diagnosis of alcoholism from comparison groups with very little overlap, and secondary sources have long quoted sensitivity around 98 per cent and specificity around 95 per cent at the original cutoff. 1 Fifty years of independent work has produced a more moderate picture. A 1983 review of the MAST literature concluded that the instrument's classification agreed with independently established diagnoses in roughly three cases out of four. 3 The largest synthesis to date, Minnich and colleagues' systematic review of 103 MAST publications in 2018, derived an aggregated internal consistency of .85 (Kuder-Richardson formula 20), judged internal and external validity to be robust, and put overall diagnostic accuracy at close to 80 per cent. 2 Reported Cronbach's alphas across individual studies fall between roughly .83 and .93, and translated versions have reported test-retest coefficients near .89.
Two patterns recur. First, performance depends heavily on the setting: the MAST does well where the base rate of established alcohol dependence is high, and its specificity falls in psychiatric and general medical inpatient populations, where lifetime adversity of the kind the items ask about is common for reasons other than alcohol. Second, the short forms lose ground - pooled internal consistency for the SMAST averages around .77 and for the bMAST around .73, both below the .80 usually asked of a screening instrument, so brevity here is bought with reliability. No study has established a minimal clinically important difference, and none could: the items ask about a lifetime, so the score is not designed to move.
KR-20 = .85
Internal consistency
Aggregated across 103 MAST publications. Minnich et al. 2018. Individual-study Cronbach's alphas range roughly .83–.93.
≈80%
Diagnostic accuracy
Overall accuracy in the 2018 systematic synthesis; an earlier review found agreement with independent diagnosis in about 3 of 4 cases.
Cutoff ≥5 (weighted 0–53)
Original validation
Selzer 1971. Secondary sources report sensitivity ≈98% and specificity ≈95%; later independent work has not reproduced figures that high.
None established
Sensitivity to change
All items are lifetime-framed. No MCID has been published and the score is not intended to move with treatment.
The shortest lifetime-framed screen. Same consequence-focused logic as the MAST in a quarter of the length; use when time is very short and you want a history.
MAST
22
5–8 min
When you want a detailed lifetime account of alcohol-related harm - in assessment, forensic or treatment-entry settings rather than routine screening.
SMAST-G / MAST-G
10 / 24
2–6 min
In older adults, where work problems and drink-driving arrests are poor markers and the standard MAST items lose sensitivity.
bMAST
10
~2 min
A shorter MAST when the full form is impractical, accepting lower internal consistency (around .73) than the parent instrument.
06 - When to use it
Right tool, wrong tool.
Reach for it when
-You need a detailed lifetime history of alcohol-related harm, not a snapshot of current intake.
-The setting has a high base rate of established alcohol dependence - addiction services, treatment intake, forensic assessment.
-You want a free, self-administered instrument with a fifty-year literature behind it.
-You are corroborating a suspected alcohol use disorder where the patient has minimised consumption but the consequences are visible.
Reach for something else when
-You are screening routinely in primary care or mental health - use the AUDIT or AUDIT-C, which detect hazardous drinking before harm.
-You want to know how much the patient is drinking now. No MAST item asks.
-You want to monitor response to treatment. The score is lifetime-based and will not fall.
-The patient is an older adult - consider the MAST-G or SMAST-G instead.
-The patient is an adolescent - the MAST is not validated in under-18s; use the CRAFFT.
-Other drugs are the concern - use the DAST-10 or the CAGE-AID.
-You are working with psychiatric or medical inpatients and need a specific result; expect a high false-positive rate here.
07 - Confidence & precision
Reading the score with care.
The MAST is internally consistent - an aggregated KR-20 of .85 across a hundred studies is respectable for a twenty-two-item binary scale - but consistency is not the same as precision, and the concept of a meaningful change score does not apply. Because every item asks about a lifetime, a MAST total is closer to a cumulative tally than to a measurement on a continuum; it can rise as harms accumulate but it cannot fall, and repeating it after an intervention is not informative. There is no published standard error of measurement and no MCID, and none should be expected. Treat the number as an index of how much damage alcohol has done, with the 3–5 band explicitly uncertain and warranting a conversation rather than a conclusion. If you need a measure that moves, use the AUDIT with its twelve-month frame, or a direct consumption record.
08 - Limitations
What it cannot tell you.
Entirely lifetime-framed. It cannot distinguish past from present problems, cannot track change, and will keep scoring positive in someone with years of stable abstinence.
No item asks about quantity or frequency, so hazardous drinking that has not yet caused visible harm is missed altogether.
Face-valid and easy to see through, which makes it vulnerable to minimisation where the respondent has something at stake - employment, custody, a licence.
Specificity falls in psychiatric and general medical inpatient populations, where the adversity the items ask about is common for reasons unrelated to alcohol.
Developed before DSM-III and framed around "alcoholism"; its content maps onto the moderate-to-severe end of DSM-5 alcohol use disorder rather than the full spectrum.
Multiple versions circulate - 25-item weighted, 24-item modified, 22-item unit-scored, SMAST, bMAST, MAST-G - with different ranges and different cutoffs. A "MAST score of 6" is meaningless without knowing which version produced it.
Poor fit for older adults and not validated in adolescents.
Largely displaced by the AUDIT in contemporary screening guidelines, so a MAST result may need translating for colleagues who expect AUDIT bands.
[4]Blow FC, Brower KJ, Schulenberg JE, Demo-Dananberg LM, Young JP, Beresford TP. The Michigan Alcoholism Screening Test - Geriatric Version (MAST-G): a new elderly-specific screening instrument. Alcoholism: Clinical and Experimental Research, 16, 372. (1992) ↩
See how Aisel removes friction where it costs most. A 20-minute walkthrough tailored to your clinic.
We value your privacy
We use cookies to analyse site usage and improve your experience. Analytics and embedded media (e.g. YouTube) only load if you accept. Read our cookie policy.