GAD-7 Anxiety Screening Score

Background.

Before anything else: if you are in crisis, or thinking about suicide or hurting yourself, get help now rather than finishing a questionnaire. In the United States, call or text 988 to reach the 988 Suicide & Crisis Lifeline — free, confidential and staffed 24 hours a day, 365 days a year — or chat at chat.988lifeline.org. Outside the United States, contact your own country's crisis line, and call your local emergency number if anyone is in immediate danger. The GAD-7 does not ask about self-harm, which is exactly why this reminder has to sit here rather than be triggered by an answer: severe anxiety and suicidal thinking co-occur often, and no seven questions about worry will surface it.

The GAD-7 is a screening questionnaire, not a diagnosis. It asks how often, over the last two weeks, seven specific problems have bothered you — feeling nervous or on edge, being unable to stop worrying, worrying too much about different things, trouble relaxing, restlessness, irritability, and a fear that something awful might happen. Each is answered on the same four-point scale, and the answers add to a whole number from 0 to 21. That number describes how many anxiety symptoms you reported and how often. It cannot tell you whether you have generalised anxiety disorder, it cannot rule it out, and it certainly cannot tell you what to do about it. Hyperthyroidism, caffeine, stimulant medication, alcohol withdrawal, arrhythmias and sleep deprivation all produce the same seven symptoms. A clinician can distinguish them. A form cannot.

The instrument was published by Robert Spitzer, Kurt Kroenke, Janet Williams and Bernd Löwe in the Archives of Internal Medicine in 2006. They developed and validated it in 2,740 adult patients drawn from 15 primary-care clinics in the United States between November 2004 and June 2005, with 965 of those patients also interviewed by a mental-health professional blinded to the questionnaire result. The companion paper from the same study, published in the Annals of Internal Medicine in 2007, found that 19.5% of those primary-care patients had at least one anxiety disorder, that impairment climbed steeply as disorders stacked up, and that 41% of patients with an anxiety disorder were receiving no treatment at all. That last figure is the reason short screeners exist.

The severity ranges on this page come from that original paper, which states that cut points of 5, 10 and 15 can be read as the lower edges of mild, moderate and severe anxiety. That gives four bands: 0-4, 5-9, 10-14 and 15-21. Read them as descriptions of a score, not of a person. A total of 11 falls in the range the authors called moderate; it does not make anyone 'moderately anxious'. The same total can be assembled very differently — seven mild symptoms or four severe ones — and a single questionnaire filled in during one difficult fortnight is a snapshot, not a diagnosis and not a trend.

The screening threshold is genuinely contested, so this page shows both candidates side by side rather than picking one quietly. The 2006 paper identified 10 as the cut point that optimised the trade-off, reporting 89% sensitivity and 82% specificity for generalised anxiety disorder. Plummer, Manea, Trepel and McMillan's independent 2016 diagnostic meta-analysis in General Hospital Psychiatry instead found a cutoff of 8 best-performing, with pooled sensitivity 0.83 and specificity 0.84, and noted that cutoffs from 7 to 10 gave similar pooled estimates. The US Preventive Services Task Force's 2023 evidence review, working at the conventional 10, reported pooled sensitivity 0.79 and specificity 0.89. So a total of 8 or 9 is a positive screen by one respectable standard and a negative screen by another — which is precisely the kind of fact that should sit next to your result, not three clicks away in an FAQ. That is why this calculator reports both comparisons as separate lines.

Knowing where a score sits in the population helps more than a label does. Löwe and colleagues administered the GAD-7 to a nationally representative German sample of 5,030 adults and reported a mean of 3.2 for women and 2.7 for men, with roughly 5% of the general population scoring 10 or more and about 1% scoring 15 or more. So a total in the high teens is genuinely unusual, while a total of 6 or 7 is common and, on its own, unremarkable — a fact that a bare severity label tends to obscure.

The eighth question, about how difficult these problems have made work, home life and getting along with other people, is part of the published instrument but is not added to the total. This calculator records it and displays it beside the score for a specific reason: symptom frequency and functional impact are different things, and the gap between them is often the most clinically interesting part of the answer. Someone reporting a total of 14 who is still working and sleeping is in a different situation from someone reporting 14 who has stopped leaving the house, and the number alone cannot tell those two apart.

One more limitation worth stating plainly. The GAD-7 was designed for generalised anxiety disorder, and it is genuinely good at that. It is a weaker instrument for panic disorder, social anxiety disorder and post-traumatic stress disorder, all of which the same study found were more or comparably common in primary care — PTSD at 8.6%, panic at 6.8%, social anxiety at 6.2%, against 7.6% for GAD itself. A low GAD-7 does not rule those out. And nothing you enter here is stored, transmitted or compared with anyone: the arithmetic runs in your browser and the page keeps no record of it.

What is gad-7 calculator?

The Generalised Anxiety Disorder 7-item scale is a short self-report screening questionnaire for anxiety symptoms. Seven items, each describing one common feature of anxiety, ask how often that problem has bothered the respondent over the previous two weeks. The four answers score 0 for 'not at all', 1 for 'several days', 2 for 'more than half the days' and 3 for 'nearly every day', giving a total between 0 and 21. A further question about functional difficulty is part of the instrument but is not added to the total. The GAD-7 is short enough to complete in about a minute, free to use, and sensitive to change over time, which is why it appears so widely in primary care, in research and in routine outcome monitoring.

What it is not is a diagnostic test. A screening instrument is built on purpose to over-refer — it is tuned to miss few people who do have the condition, which necessarily means flagging many who do not. A score above the threshold means 'this deserves a proper assessment' and nothing more precise. It was validated against a structured psychiatric interview conducted by a blinded mental-health professional, and that interview, not the questionnaire, is what actually establishes a diagnosis. The instrument also carries the boundaries of the population it was built in: adults attending general practices in the United States, answering in English. Its behaviour in other languages, cultures, age groups and clinical settings is a separate empirical question with its own literature.

How to use this calculator.

1. 01Answer each of the seven items for the last two weeks — not for today alone, and not for the past year.
2. 02Use the four options as the instrument defines them: not at all (0), several days (1), more than half the days (2), nearly every day (3).
3. 03Answer the eighth question about how difficult these problems have made daily life. It is part of the instrument but is deliberately not added to the 0-21 total.
4. 04Read the total alongside the band, both cut-point comparisons and the difficulty answer. If the two cut-point lines disagree, your total is 8 or 9 and sits in the window where published sources genuinely differ.
5. 05Compare the total with population norms rather than with a label: roughly 5% of a general adult population score 10 or more, and about 1% score 15 or more.
6. 06Take the result to a clinician. A score is the start of a conversation, not a diagnosis, and never a treatment decision.
7. 07If you are in crisis at any point, stop and call or text 988 in the United States, or your local crisis line elsewhere.

The formula.

The arithmetic is a plain sum of seven whole numbers, each between 0 and 3, so the total is an integer between 0 and 21. Nothing is weighted, nothing is reverse-scored, and no adjustment is made for age, sex or anything else — every item counts the same. The functional-difficulty question is recorded separately and never enters the total, exactly as the published instrument specifies.

Because every input is an integer, there is no rounding at any stage, intermediate or final. That is worth saying explicitly: a total cannot drift across a band boundary through accumulated floating-point error, and there is no borderline 'rounds to 10' case to argue about. A total is either 9 or it is 10. Band boundaries are applied with inclusive lower bounds — 5 and above is the mild range, 10 and above the moderate range, 15 and above the severe range — which is exactly how the 2006 paper stated them when it described cut points of 5, 10 and 15 as representing mild, moderate and severe levels of anxiety.

The two screening thresholds are separate conventions layered on the same score, and this calculator deliberately reports both rather than choosing. The 2006 validation study identified 10 as optimal, with 89% sensitivity and 82% specificity for generalised anxiety disorder against a blinded structured interview in 965 patients. Plummer and colleagues' independent 2016 meta-analysis found a cutoff of 8 best-performing, pooling sensitivity 0.83 (95% CI 0.71-0.91) and specificity 0.84 (95% CI 0.70-0.92), and reported that cutoffs of 7 through 10 produced similar pooled estimates. The USPSTF's 2023 evidence review, evaluating the conventional cutoff of 10, pooled sensitivity 0.79 (95% CI 0.65-0.94) and specificity 0.89 (95% CI 0.83-0.94). None of these numbers is wrong; they come from different study sets, different reference standards and different populations. A threshold is a policy choice about how many false positives are worth accepting to miss fewer true cases, and that choice is not the same in a specialist clinic, a general practice and an anonymous web page. Showing both comparisons is the only honest way to present a score of 8 or 9.

A worked example.

Frequently asked questions.

References& sources.

1. [1]Spitzer RL, Kroenke K, Williams JBW, Löwe B. "A brief measure for assessing generalized anxiety disorder: the GAD-7." Archives of Internal Medicine. 2006;166(10):1092-1097. doi:10.1001/archinte.166.10.1092. Original validation paper — source of the 0-3 item coding, the 0-21 range, the cut points of 5/10/15, and the 89%/82% figures at a cut point of 10. Full text retrieved 29 July 2026.
2. [2]Plummer F, Manea L, Trepel D, McMillan D. "Screening for anxiety disorders with the GAD-7 and GAD-2: a systematic review and diagnostic metaanalysis." General Hospital Psychiatry. 2016;39:24-31. doi:10.1016/j.genhosppsych.2015.11.005. Independent meta-analysis; at a cutoff of 8, pooled sensitivity 0.83 (95% CI 0.71-0.91) and specificity 0.84 (95% CI 0.70-0.92), with cutoffs 7-10 giving similar estimates. Abstract open on PubMed (PMID 26719105); full text paywalled. Retrieved 29 July 2026.
3. [3]US Preventive Services Task Force. "Anxiety Disorders in Adults: Screening." Final Recommendation Statement, 20 June 2023. Grade B recommendation to screen adults for anxiety disorders; GAD-7 pooled sensitivity 0.79 (95% CI 0.65-0.94) and specificity 0.89 (95% CI 0.83-0.94) at a cutoff of 10 or greater. Retrieved 29 July 2026.
4. [4]Löwe B, Decker O, Müller S, Brähler E, Schellberg D, Herzog W, Herzberg PY. "Validation and standardization of the Generalized Anxiety Disorder Screener (GAD-7) in the general population." Medical Care. 2008;46(3):266-274. doi:10.1097/MLR.0b013e318160d093. Source of the population norms used on this page: n = 5,030 German adults, mean 3.2 (women) and 2.7 (men), approximately 5% scoring 10 or more and 1% scoring 15 or more, alpha 0.89. Abstract open on PubMed (PMID 18388841); full text paywalled. Retrieved 29 July 2026.
5. [5]Kroenke K, Spitzer RL, Williams JBW, Monahan PO, Löwe B. "Anxiety disorders in primary care: prevalence, impairment, comorbidity, and detection." Annals of Internal Medicine. 2007;146(5):317-325. Companion paper to the GAD-7 validation study: 19.5% of 965 primary-care patients had at least one anxiety disorder (PTSD 8.6%, GAD 7.6%, panic 6.8%, social anxiety 6.2%), and 41% of those with an anxiety disorder reported no current treatment. Abstract open on PubMed (PMID 17339617). Retrieved 29 July 2026.
6. [6]988 Suicide & Crisis Lifeline (United States). "The 988 Lifeline is available 24/7/365. Your conversations are free and confidential." Call or text 988; chat at chat.988lifeline.org. Retrieved 29 July 2026.