Autism tests › Adult tests (AQ & RAADS-R)
Autism tests for adults
Many parents researching their child recognize themselves along the way — and many adults come to the question on their own. Here's what the adult self-report questionnaires actually measure, and what a real adult assessment looks like.
AQ — Autism-Spectrum Quotient
| Type | Self-report screening questionnaire |
|---|---|
| Format | 50 statements rated agree/disagree, ~10 minutes |
| Origin | Developed by Simon Baron-Cohen's group at Cambridge; freely available for personal use |
| Reading a score | 32 or more was the original threshold in Baron-Cohen's 2001 paper, where 80% of autistic adults but only 2% of controls reached it. Later work proposed lower clinical cut-offs — 26 (Woodbury-Smith, 2005) and 29 (Broadbent, 2013). It measures traits, not diagnosis. |
The AQ asks about social preferences, attention switching, detail focus, communication, and imagination. It's a reasonable first data point — and also famously blunt: high scores occur without autism, and autistic adults who mask heavily can score low.
RAADS-R — Ritvo Autism Asperger Diagnostic Scale, Revised
| Type | Self-report screening questionnaire (designed for clinical settings) |
|---|---|
| Format | 80 statements rated on a 4-point scale, including "true now / true only when I was younger" — capturing developmental history |
| Structure | Four subscales: language, social relatedness, sensory-motor, and circumscribed interests |
| Reading a score | 65 or above was the threshold in the 2011 validation study — but see the caveat below before reading anything into your number |
The RAADS-R was designed to help identify adults whose autism was missed in childhood, and its developmental-history angle is its strength. It was validated for use with a clinician present — the free online versions circulating everywhere weren't part of that validation, so treat online scores as a structured way to organize your observations, not a result.
The cutoff deserves a warning of its own. The original study reported near-perfect accuracy at a score of 65, but it compared autistic adults against a general comparison group. In real clinical settings — where the people taking it are, by definition, those already wondering — that threshold turns out to be very loose, and subsequent research has proposed substantially higher cut-points. Practically: crossing 65 on an online version tells you far less than the number's precision suggests. It is a reason to seek assessment, nothing more.
An online score is a conversation starter, not an identity document — in both directions. A high score doesn't make you autistic; an unremarkable score doesn't un-make you. If the question matters to your life — work, relationships, mental-health care, self-understanding — the score's job is done the moment it convinces you to seek a real assessment.
What a real adult assessment involves
- A detailed clinical interview covering your current life and your childhood (school reports, parents' or siblings' memories help enormously if available).
- Often a structured instrument — the ADOS-2 Module 4 (for fluent adults) and/or standardized questionnaires like the SRS-2 adult forms.
- Ruling in/out overlapping explanations: ADHD, anxiety, OCD, trauma — which frequently co-occur with autism rather than excluding it.
- Judgment against DSM-5-TR / ICD-11 criteria, including the requirement that traits were present (even if unrecognized or masked) since early development.
Assessment quality for adults varies widely — look for clinicians who explicitly assess adults, mention masking and late-diagnosed presentations, and welcome the developmental-history work. Autistic-led communities (see resources) maintain good guidance on finding assessors.
Medical disclaimer: Educational information only. The AQ and RAADS-R belong to their respective authors. This page cannot diagnose anyone; consult a qualified clinician.