Eating disorder test (EAT-26)
See what the EAT-26 covers, how its standard cut-off works, and why some behavioural answers prompt a referral regardless of the total. This is a sensitive subject, so the page keeps a gentle, supportive tone.
The EAT-26 (Eating Attitudes Test-26) is a short, widely used questionnaire about attitudes and behaviours around eating, weight and food. It was developed by Garner and colleagues in 1982 and is the most commonly used eating-disorder screening tool in the world. It takes about five minutes to answer. A higher result does not define a person; concerns about eating are common and support is available and effective.
The model
What it measures
The EAT-26 covers one broad area: disordered-eating attitudes and behaviours - things like preoccupation with food, fear of weight gain, dieting and a sense that food or eating feels controlling. Its twenty-six items ask how often each pattern is present, and the answers are summed into a single total.
The items group into three themes - dieting, bulimia and food preoccupation, and oral control - but for screening the EAT-26 produces one total. A higher total means more of these attitudes are present and more often. Alongside the scored items, a short set of behavioural questions (for example about bingeing, vomiting or recent weight change) are read separately as risk indicators, because some of these can warrant a referral on their own.
- EDisordered-eating attitudes
How often attitudes and behaviours linked to disordered eating - preoccupation, fear of weight gain, dieting and control - are present.
Facets: Preoccupation with food, Fear of being overweight, Dieting and food restriction, Feeling that food is controlling, Bingeing and loss of control, Guilt after eating, Desire to be thinner, Oral control and self-monitoring.
The evidence
Science and validity
The EAT-26 is the most established self-report eating-disorder screener and is often the benchmark against which newer measures are compared. Internal consistency is strong - Cronbach's alpha is typically around .90 - and it distinguishes people with eating disorders from comparison groups well. It is widely used in epidemiology and in school, university and clinical screening across dozens of languages.
Each of the twenty-six items is rated on a six-point frequency scale, and the scale is asymmetric: only the three most symptomatic answers count (scored 3, 2 and 1, with item 26 reverse-keyed), so the total ranges from 0 to 78. There are no population percentiles. A total of 20 or above flags a level of concern that warrants a fuller, compassionate evaluation with a qualified professional. Importantly, the behavioural questions can flag risk independently: clear answers about bingeing, vomiting, laxative use or significant recent weight loss warrant a referral even when the total sits below 20.
How it is scored
Score bands and what they mean
This is a screening questionnaire. The total is read against established cut-off bands, not a population percentile and not a diagnosis - the bands flag how strongly recent symptoms are showing up, and where it may help to talk to someone.
- 0-19Below the cut-offLower concern on this screen; if you have worries about eating, a professional can still help and support works.
- 20-78At or above the cut-offA level of concern that warrants a fuller, compassionate evaluation with a qualified professional.
- Behavioural red flagsRefer regardless of totalClear answers about bingeing, vomiting, laxative use or significant recent weight loss warrant a referral whatever the total.
How it works
What the questions feel like
Illustrative statements showing the style of the items. These are examples, not the official scored items.
How often are you preoccupied with thoughts about food?
Illustrative example in the style of the screener, not the official scored item.
How often do you feel very uneasy at the thought of gaining weight?
Illustrative example, not the official scored item.
How often do you feel that food has too much control over your life?
Illustrative example, not the official scored item.
How often do you feel guilty after eating?
Illustrative example, not the official scored item.
Honest strengths and limitations
Strengths
- The most widely used and best-validated eating-disorder screener in the world, in 30+ languages.
- Short (about five minutes) and free for individual and research use.
- A clear cut-off plus behavioural red-flag questions, so risk can be caught even when the total is low.
Limitations
- It is a screening questionnaire, not a diagnosis - a higher total signals that a caring conversation with a professional may help, not that an eating disorder is present.
- A total below the cut-off does not rule concerns out, and a higher total does not confirm anything; the behavioural questions matter alongside the score.
- Like all self-reports it reflects how someone feels right now and can be shaped by recent events and how the questions are read.
Checking in on how you are doing?
Screeners like this are informational, not a diagnosis. The free Snapshot is a private, structured way to check in on how you have been feeling lately.
Frequently asked questions
What does the EAT-26 measure?
It covers attitudes and behaviours linked to disordered eating - preoccupation with food, fear of weight gain, dieting, and a sense that food or eating feels controlling. Its twenty-six items are summed into a single total, and a few behavioural questions are read separately as risk indicators.
How is the EAT-26 scored?
Each item uses a six-point frequency scale that is scored asymmetrically - only the three most symptomatic answers count (3, 2 and 1, with one item reverse-keyed) - for a total of 0 to 78. The total is read against a standard cut-off of 20 rather than a percentile. A total of 20 or above flags concern, and the behavioural questions can warrant a referral on their own.
Is the EAT-26 a diagnosis?
No. The EAT-26 is a screening questionnaire, not a diagnosis. A higher total means it may help to speak with a qualified professional, who can look at the full picture with care. Eating difficulties are common and treatable. If you are struggling or in crisis, please contact a qualified professional or a local eating-disorder helpline or crisis service - many countries have a free, confidential eating-disorder helpline by phone, text or webchat.
Can I take the EAT-26 on Psychology.me?
This page is informational - we do not offer the EAT-26 itself. If you would like a private, gentle way to check in on how you have been doing, the free wellbeing Snapshot is a supportive place to start.
Related tests
- Garner, D. M., Olmsted, M. P., Bohr, Y., & Garfinkel, P. E. (1982). The Eating Attitudes Test: psychometric features and clinical correlates. Psychological Medicine, 12(4), 871-878.
- Garner, D. M., & Garfinkel, P. E. (1979). The Eating Attitudes Test: an index of the symptoms of anorexia nervosa. Psychological Medicine, 9(2), 273-279.
The EAT-26 was developed by David M. Garner and colleagues (1982); it is free for individual and research use with attribution, while commercial web use requires a licence from the rights holder. This independent informational page describes the instrument and does not reproduce its scored items.