AQ-10

What the AQ-10 is

The AQ-10 is a ten-question autism screener, published by Allison, Auyeung and Baron-Cohen in 2012 at the Autism Research Centre in Cambridge. It is used across the NHS as an early step on the adult autism pathway, which is the main reason it is on this page. Many people meet it first in a GP surgery.

It is a short form. The full Autism-Spectrum Quotient has fifty questions, and the ten here were selected because they separated autistic from non-autistic respondents more sharply than the rest. That makes the AQ-10 quick and quite good at sorting a large group. It also means it is a sample of autistic experience rather than a description of it.

The ten questions cover four areas. Noticing detail and small sounds that other people filter out. Moving attention from one thing to another, including how easily you pick up a task after an interruption. Reading intention and feeling in other people, from a face or from a story. Collecting and organising information about categories of things.

Four questions score when you agree with them and six score when you disagree, which is a design choice rather than an accident. Mixing the direction stops the questionnaire rewarding a habit of agreeing with everything, and it is one reason the instruction is to answer instinctively rather than to work each item out.

How your score was worked out

Each of the ten questions scores one point when you answer it in the direction associated with autistic experience. For four of them that direction is agreeing. For the other six it is disagreeing. The total runs from 0 to 10.

Strength of agreement does not change the score. Definitely Agree and Slightly Agree count the same, and so do Definitely Disagree and Slightly Disagree. The published guidance is that a score of 6 or more means a full autism assessment is worth considering.

That cutoff is a sorting line, not a verdict. A 5 and a 6 describe almost the same set of answers. If you scored close to the line in either direction, read your answers again rather than reading the number.

What a score of 6 or more means

It means enough of your experience overlaps with autistic experience that an unhurried look by a clinician would give you a real answer. It is not a diagnosis and it cannot be one. Ten questions cannot examine a developmental history.

An autism assessment looks at how you have communicated and related to other people since childhood, your sensory experience, how you think and process information and what all of that has cost or given you across your life. It is mapped to DSM-5-TR and ICD-11 in line with NICE guidelines, and it usually includes information from someone who knew you when you were young.

It also looks at what else is happening. ADHD and autism often occur in the same person and each can hide the other, so a clinician who assesses only the one you asked about can give you half an answer. Anxiety and depression frequently occur alongside both, sometimes because a person has spent years working very hard to look ordinary. A good assessment sorts out what is causing what rather than stopping at the first thing it finds.

What a score below 6 means

Less than people generally assume. A short screener with ten items has to be blunt, and being blunt is how it stays quick.

Nothing on the AQ-10 asks whether you have learned to do socially what other people do instinctively. Nothing asks what it costs you to hold a conversation, how long you need alone afterwards or whether you rehearse what you are going to say before you say it. Someone who has spent twenty years studying how other people talk can answer the question about reading between the lines with an honest disagreement, because by now they can do it, having learned it as a skill rather than by instinct.

The same is true of the detail items. A person who has always noticed every small sound in a room may not think of that as noticing anything unusual, because it is simply what a room sounds like to them. Self-report only works where you have something to compare yourself against, and nobody gets to hear a room the way somebody else hears it.

So a result below 6 rules very little out. If the questions felt like they were describing your life even though the total was low, that reaction is better evidence than the total.

What the AQ-10 was built around

The AQ family was written against the description of autism available at the time, and that description was drawn mainly from people who were identified as children, mostly boys, and mostly because their differences were visible to adults around them.

It follows that the items ask about the things that were visible. There is nothing on the AQ-10 about camouflage, about exhaustion after social contact, about the difference between how you are at work and how you are at home or about the difference between what you can do and what it costs you to do it. Those are the features that most often describe autistic women, non-binary people and trans people, and they are the reason a great many adults reach a first diagnosis late.

This is not a flaw the authors hid. It is what a ten-item screen from a particular period of research can and cannot cover, and it is worth knowing before you treat a number out of ten as your answer.

If the result does not match your experience

Two other free screeners here were designed for the questions the AQ-10 does not ask.

MASC asks about camouflaging directly. Rehearsing conversations before you have them. Monitoring your face and your posture while you speak. Copying other people so you come across the way they expect. The hours or days of recovery afterwards. It was written for the presentations that are missed most often, which are women, non-binary people and trans people, and it was validated on an initial cohort of women and non-binary adults. The supplemental questions about attention and recovery after forced focus are still being validated and are reported separately.

The AQ-50, the full fifty-item version of the same family, gives a longer read with five separate subscales instead of one total. Where your subscale pattern is uneven, that pattern is often more useful than any single figure.

Neither is presented as better than the AQ-10. They are designed for different questions and each states its validation status on its own page. If two screeners disagree about you, bring both results to your assessment.

How to use this result

As one piece of information, dated today, that you can put next to everything else you know about yourself.

If you decide to go further, a few things help. Write down the examples that came to mind while you were answering, because those specifics are what a clinician can work with and they are easy to forget on the day. School reports are useful if you still have them. So is a conversation with a parent, sibling or old friend about what you were like as a child, with one caution worth holding: the people around you saw the version of you they were shown, so their memory of your childhood may be a memory of your masking.

Our assessments are online, no GP referral is required and the written report arrives within six weeks of your first appointment.

Common questions

Is the AQ-10 an autism diagnosis?

No. It is a ten-question screener used as an early step on the NHS adult autism pathway. A score of 6 or more is the published signal that a full assessment is worth considering. A diagnosis needs a clinical assessment of your developmental history, your communication and relationships, your sensory experience and how you process information, mapped to DSM-5-TR and ICD-11.

Why is the cutoff 6 out of 10?

That threshold comes from the 2012 publication by Allison, Auyeung and Baron-Cohen, where 6 gave the most useful separation between autistic and non-autistic respondents. It is a sorting line rather than a boundary in nature, which is why a 5 and a 6 tell you almost the same thing.

Can I be autistic and score below 6?

Yes. The AQ-10 asks nothing about masking, about exhaustion after social contact or about social skills you learned deliberately rather than by instinct. If you have spent years studying how other people interact, you can answer several items with an honest disagreement and still be autistic. If the questions described your life despite the total, that reaction is worth more than the number.

Should I take the AQ-50 as well?

It is often worth it. The AQ-50 is the full version of the same family and it reports five subscales rather than a single total, so an uneven pattern becomes visible. It takes about ten minutes. If you think camouflaging is a large part of your picture, MASC asks about that directly.

Does the NHS or the HSE accept an AQ-10 result?

As a screening step, yes, and it is widely used that way. It is not accepted as a diagnosis by anyone, because it is not one. What is accepted is a diagnostic report from a qualified clinician written in line with NICE guidelines, which is what our assessments produce within six weeks of your first appointment.