What happens in an adult autism assessment?

What the assessment is trying to establish

It helps to know the question the clinician is answering, because it isn’t “are you autistic enough.”

They’re testing whether your history and current life meet the diagnostic criteria: persistent differences in social communication and interaction, alongside restricted or repetitive patterns of behaviour, interests, or activities. The DSM-5-TR, the American Psychiatric Association’s diagnostic manual, says traits must be present in the early developmental period, while allowing that they may be masked by learned strategies or may not become fully apparent until social demands exceed a person’s capacities (American Psychiatric Association, 2022).

That last clause matters more than any other sentence in this page. The criteria explicitly anticipate someone who coped for decades and then didn’t.

The second job is differential: working out what else could account for what you’re describing, and what might be going on alongside it. Published guidance asks clinicians to assess for other neurodevelopmental conditions, mental health conditions, and physical or communication difficulties as part of the same process (NICE, 2021). That’s why you’ll be asked about things that feel unrelated.

Before the appointment

Some services send questionnaires ahead of time. If yours does, filling them in is part of the information-gathering process.

These are screening or structuring instruments, not tests you pass. NICE suggests considering the ten-item Autism-Spectrum Quotient for initial screening and lists the Adult Asperger Assessment and Ritvo Autism Asperger Diagnostic Scale-Revised among several tools that may aid a more complex assessment (NICE, 2021). A questionnaire score doesn’t diagnose you or rule autism out by itself. It gives the clinician one source of information to consider.

Two things are worth doing while you fill them in. Answer as accurately as you can, including what happens when you use coping strategies and what those strategies cost in effort or recovery. And keep notes on the examples that come to mind, because in the room you may be asked for specifics and they’re much harder to retrieve under pressure.

What happens in the room

A detailed conversation is a common part of an adult assessment. It often covers two broad areas.

Your developmental history. What you were like as a child: how you played, how you got on with other children, what you were fixated on, how you handled change, what your family remembers. If this feels like an odd amount of attention on a childhood you barely remember, that’s because the criteria are anchored there.

How life works now. Friendships and relationships, work or study, sensory experience, routines, what happens when plans change, what you do to recover from social contact, and what you find yourself doing deliberately to appear at ease. Expect questions about the effort as well as the outcome.

Some services add a structured observation, such as the ADOS-2. It’s a session of conversation and activities designed to give the clinician standardised things to watch—how you tell a story, how you handle an open-ended prompt, what you do with an object. It can feel strange and a little staged, because it is. Not every assessment includes one; NICE lists an ADOS observation tool among several options for structuring a complex assessment (NICE, 2021).

Whether you need someone who knew you as a child

NICE says a comprehensive assessment should, where possible, involve a family member, partner, carer, or other informant, or use documentary evidence such as school reports of current and past behaviour and early development.

If you don’t have an informant or documents, tell the service when you book and ask how it approaches developmental history in that situation. Informants don’t have to be parents. We cover ways to look for documentary history in more depth on whether you need childhood records, which is written for ADHD; your autism-assessment provider can tell you what it accepts.

What the tools can and can’t tell you

Here’s the part most pages skip, and it’s useful to know walking in: a structured observation can add information, but it isn’t a standalone diagnosis, and its performance depends on the population and clinical setting in which it is used.

One retrospective study reviewed 88 adults referred in 2017 and 2018 to a single specialist adult ADHD and autism service in the NHS, the UK’s public health service, in South and West Yorkshire. The service assessed adults without intellectual disability, and the study required an IQ above 70. Complete ADOS-2 Module 4 data were available for 83 people, and 26 of the 88 received an autism diagnosis through multidisciplinary clinical consensus. In that specific sample, the ADOS-2 threshold had 92% sensitivity and 57% specificity. Those figures should not be treated as universal performance estimates; the authors recommended caution when interpreting an ADOS-2 score on its own (Adamou, Jones and Wetherhill, 2021).

Masking and compensatory strategies can limit what is visible in a brief observation. DSM-5-TR explicitly allows for traits being masked by learned strategies later in life. That is a reason to discuss strategies, effort, and experiences that may not be observable in the room, not evidence for a particular accuracy rate in any subgroup.

Two practical consequences follow.

The first is that a comprehensive assessment doesn’t rest on one instrument. History, present-day detail, and observation can be considered alongside questionnaires and, where possible, an informant account or documentary evidence.

The second is to answer accurately instead of trying to act more or less autistic. A clinician can observe that you made eye contact. They can’t observe that you were counting seconds to know when to break it. Explain learned strategies, the effort they take, and what happens before or after the appointment when that context matters, even when it feels like over-explaining.

How it ends

How feedback is delivered and what documentation you receive vary by provider. NICE says the service should discuss at the beginning how the outcome will be fed back and should individualise that feedback.

Some providers give a written report, some provide a summary letter, and documentation may be handled differently elsewhere. Ask at the start what you’ll receive, how long it usually takes, and whether it will meet any documentation requirements you have for work, education, benefits, or healthcare.

Outcomes come in more than two shapes. A diagnosis, no diagnosis, or an inconclusive result with a recommendation for further assessment are all possible, as is a diagnosis of something else entirely. If the answer is no and it doesn’t sit right, that’s information and not a verdict—a second opinion is a legitimate next step, and clinicians vary in how familiar they are with adult presentation.

Waiting may take much longer than the appointments themselves. A 2025 retrospective review covered 408 completed neurodevelopmental assessments conducted between October 2021 and May 2022 by 30 teams across 10 of Scotland’s 14 health boards. Among 202 completed cases from 12 adult teams, the median time from referral to the written outcome was 252 days, with a wide interquartile range of 106 to 611 days. Cases ultimately diagnosed with ADHD without autism waited longer than cases diagnosed with autism without ADHD (Maciver et al., 2025). This was a mixed sample of completed autism, ADHD, and other neurodevelopmental assessments in one national system and time period, not a universal or current autism wait estimate. Ask your provider what its present pathway looks like.

If you haven’t decided whether to start, whether assessment is worth pursuing is the question that comes first, and self-identification is a reasonable place to land if the doors a diagnosis opens aren’t ones you need. To get your own history in order beforehand, the NeuroDiversion AuDHD self-reflection is built for that step. Services vary, so treat the above as the shape of the thing and ask yours what its own process looks like.