The Free RAADS-R Test: Take it Online & How It Works
A plain-English guide to the Ritvo Autism Asperger Diagnostic Scale, Revised. What it measures, how it is scored, how long it takes, whether it is free, and what the independent research actually found when it was tested in real clinics.
- What the RAADS-R test is
- How the RAADS-R test works
- How many questions and how long
- How the RAADS-R test is scored
- How accurate the RAADS-R test is
- Why the 65 cutoff is disputed
- Is the RAADS-R test free?
- RAADS-R versus AQ and other tests
- Who the RAADS-R test is for
- Limitations worth knowing
- Frequently asked questions
- Research and sources
What the RAADS-R test is
The RAADS-R test is an 80-item self-report questionnaire designed to help identify autism in adults, particularly those who reached adulthood without being diagnosed. Its full name is the Ritvo Autism Asperger Diagnostic Scale, Revised. It was published in 2011 by Riva Ariella Ritvo, Edward Ritvo and colleagues in the Journal of Autism and Developmental Disorders.
It was built to solve a specific problem. Most autism instruments in use at the time had been developed from observations of children, which made them a poor fit for a forty-year-old who had spent decades developing workarounds. The RAADS-R test was written for exactly that person, and it is still one of the two most widely used adult self-report measures alongside the Autism Spectrum Quotient.
One design feature sets it apart from almost every alternative. The RAADS-R test asks you to consider your whole life rather than only the present. That matters because autism is developmental by definition, present from early childhood even when nobody noticed it, so a questionnaire that only asks about this month can miss someone entirely.
How the RAADS-R test works
You read 80 statements about social situations, sensory experience, language and interests, and for each one you choose from four options describing when it has been true for you:
- True now and when I was young
- True only now
- True only when I was younger than 16
- Never true
This lifetime structure is the distinctive part of the RAADS-R test. A statement that was true in childhood but is no longer true still contributes to your score, because the instrument is looking for a developmental pattern rather than a snapshot of how you are this week.
The items are grouped into four areas:
| Subscale | What it covers |
|---|---|
| Social relatedness | The largest section. Reading intent, managing conversation, friendship, social reciprocity. |
| Circumscribed interests | Depth and intensity of interests, attention to detail, routines. |
| Sensory-motor | Reactions to sound, light, touch and texture, plus repetitive movement. |
| Language | Literal interpretation, figures of speech, pragmatic use of language. |
Worth knowing: a 2024 analysis found the RAADS-R test functions statistically as a single dimension rather than four independent ones. The subscales remain useful as descriptive categories, but they are not measuring four separate things in the way the structure implies.
How many questions and how long it takes
The RAADS-R test has 80 questions. Most people complete it in 15 to 20 minutes, though it can take longer if you think carefully about the lifetime framing, which you should.
That length is a genuine drawback in clinical settings, and researchers have said so. It is why shorter derivatives exist. The RAADS-14 Screen reduces it to 14 items for use in busy psychiatric outpatient services, reaching sensitivity of 97 per cent but specificity of only 46 to 64 per cent at its threshold of 14.
How the RAADS-R test is scored
Each answer is worth 0 to 3 points, and the total runs from 0 to 240. Some items are reverse-scored so that agreeing with everything does not automatically produce a high result.
The original paper set the threshold at 65. Scoring above 65 was described as consistent with autism; scoring below it was described as making autism unlikely.
| Score | Original interpretation | Revised interpretation proposed in later work |
|---|---|---|
| Below 65 | Autism unlikely | Low level of autistic traits |
| 65 to 105 | Above threshold | Some autistic traits |
| 106 to 139 | Above threshold | Traits consistent with autism |
| 140 and above | Above threshold | Pronounced autistic traits |
The revised bands exist because the single cutoff of 65 turned out to be close to useless in clinical populations. Those updated thresholds report sensitivity and specificity of roughly 81 per cent, which is considerably more honest than the original claim.
How accurate the RAADS-R test is
This is the question most pages about the RAADS-R test answer badly, usually by quoting the 2011 paper and stopping there. Here is the full picture.
What the original study found
Ritvo and colleagues tested 201 autistic adults against 578 comparison participants, including both non-autistic controls and people with other psychiatric diagnoses. At a cutoff of 65 they reported sensitivity of 97 per cent and specificity of 100 per cent, with test-retest reliability of 0.987. Those are extraordinary figures, and they are the ones still quoted almost everywhere.
What independent studies found afterwards
The picture changed substantially once other researchers tested the RAADS-R test outside the original setting.
A 2021 UK study by Jones and colleagues, published in Autism Research and Treatment, gave the RAADS-R test to 50 adults on the waiting list of a specialist autism service. Sensitivity was 100 per cent, but specificity was 3.03 per cent. Ninety-eight per cent of participants scored above 65, while only 34 per cent went on to receive a diagnosis. The area under the curve was 0.45, which is no better than chance. The authors concluded that the RAADS-R test “lacks predictive validity and is not a suitable screening tool for adults awaiting autism assessments”.
Other work points the same way. Conner and colleagues assessed 93 adults at a specialist service and found no difference in RAADS-R scores between those who received a diagnosis and those who did not. A 2024 study found that over half of general mental health clinic patients scored above the threshold. Pooled estimates put sensitivity around 0.75 and specificity around 0.71, which is fair at best.
| Study | Setting | Sensitivity | Specificity |
|---|---|---|---|
| Ritvo et al. 2011 (original) | Research sample, clinician present | 97% | 100% |
| Jones et al. 2021 | UK specialist autism service | 100% | 3% |
| Pooled estimates | Mixed clinical samples | ~75% | ~71% |
| German version (RADS-R) 2024 | Diagnosed vs control sample | 92.5% | 93.6% |
| Revised thresholds | Clinical populations | ~81% | ~81% |
Why the gap is so large
Three reasons, and none of them mean the original researchers did anything wrong.
Who you compare against changes everything. Separating autistic adults from the general population is easy. Separating them from other people who have already decided they might be autistic and joined a waiting list is extremely hard, because that group self-selects for exactly these traits.
Administration conditions differed. In the original study a clinician was present. In most real use, including every online version, people complete it alone.
Overlapping conditions inflate scores. ADHD, social anxiety, complex trauma and burnout all push scores up. The RAADS-R test cannot separate them, which is why it performs so poorly in psychiatric settings where those conditions are common.
Why the 65 cutoff is disputed
The practical problem with a threshold of 65 is that almost everyone who seeks out the RAADS-R test clears it. If 98 per cent of a waiting list scores above the line and only a third are diagnosed, the line is not sorting anyone.
That is why the revised bands were proposed. They do not replace the original score, they reinterpret it: 65 stops being a verdict and becomes the bottom of a range, with meaningfully different interpretations at 106 and 140.
If you have taken the RAADS-R test and scored 70, the useful reading is “some autistic traits reported”, not “above the autism threshold”. The difference matters.
Is the RAADS-R test free?
Yes, for personal, clinical and research use. The original 2011 paper was published under a Creative Commons Attribution Non-Commercial licence, which permits free use and reproduction provided the use is not commercial and the authors are credited.
That non-commercial clause is more restrictive than many people realise. It means the RAADS-R test can be used freely by individuals, clinicians and researchers, but websites that run advertising or charge for access are outside the terms of the licence. A good many sites hosting it alongside advertising are technically in breach of that clause.
RAADS-R test versus the AQ and other instruments
| RAADS-R | AQ (Autism Spectrum Quotient) | AQ-10 | |
|---|---|---|---|
| Items | 80 | 50 | 10 |
| Time | 15 to 20 minutes | 10 to 15 minutes | 2 minutes |
| Timeframe | Whole life | Present day | Present day |
| Score range | 0 to 240 | 0 to 50 | 0 to 10 |
| Threshold | 65 (disputed) | 32 (also debated) | 6 |
| Asks about masking | No | No | No |
| Typical use | Adults undiagnosed into adulthood | Research, trait measurement | Brief referral screen |
In head-to-head comparisons the RAADS-R test and the AQ perform similarly, which is to say both are fair rather than good at predicting diagnosis. Neither asks about camouflaging, which is the main reason both under-detect women and late-diagnosed adults. The CAT-Q covers that gap but measures masking specifically rather than autism.
Who the RAADS-R test is for
It was designed for adults aged 18 and over who were not diagnosed in childhood. It is used in research, by some clinicians as a structured starting point for conversation, and very widely by people investigating the question for themselves.
It is not designed for children or young teenagers, both because of the reading level and because the lifetime framing assumes a long life to look back across. It is also not designed to be used alone as a basis for any decision.
Limitations worth knowing before you take it
It cannot diagnose. No self-report questionnaire can, and the RAADS-R test is explicit about this in its own documentation.
It does not ask about masking. Someone who has camouflaged successfully for thirty years may answer the social items in a way that understates their experience.
It cannot separate autism from its lookalikes. This is the clearest finding from the independent research, and the reason specificity collapses in psychiatric settings.
It relies on memory of childhood. Half the response options ask what was true before you were sixteen, which some people simply cannot answer reliably.
Online versions vary. Many sites hosting the RAADS-R test score it incorrectly, omit reverse-scoring, or present the 65 threshold as a verdict. If you take it online, check the source.
Frequently asked questions about the RAADS-R test
What is the RAADS-R test?
The RAADS-R test is an 80-item self-report questionnaire that measures autistic traits in adults. Published in 2011 by Ritvo and colleagues, it was designed specifically for adults who were not diagnosed as children. Scores range from 0 to 240.
How many questions are in the RAADS-R test?
80 questions, grouped into four areas: social relatedness, circumscribed interests, sensory-motor experience, and language.
How long does the RAADS-R test take?
Between 15 and 20 minutes for most people. There is no time limit, and taking longer is fine because several questions ask you to recall what was true before the age of 16.
Is the RAADS-R test free?
Yes, for personal, clinical and research use. It is published under a Creative Commons Non-Commercial licence, which permits free use with attribution but does not permit commercial use.
What is a high score on the RAADS-R test?
The original threshold was 65 out of 240. Later work proposed more useful bands: below 65 is a low level of traits, 65 to 105 indicates some autistic traits, 106 to 139 indicates traits consistent with autism, and 140 or above indicates pronounced traits.
How accurate is the RAADS-R test?
The original study reported 97 per cent sensitivity and 100 per cent specificity. Independent studies found much weaker performance, including one UK clinical study reporting 3 per cent specificity. Pooled estimates put both sensitivity and specificity in the 70 to 80 per cent range. It is better understood as a structured description of traits than as a predictor of diagnosis.
Can the RAADS-R test diagnose autism?
No. It is a screening questionnaire. Diagnosis requires a clinical assessment including developmental history and, usually, several hours with a qualified professional.
Does a high RAADS-R score mean I am autistic?
Not on its own. A high score means you reported a substantial number of autistic traits. ADHD, social anxiety, complex trauma and long-term burnout also produce elevated scores, and a questionnaire cannot separate them.
Does a low RAADS-R score rule out autism?
No. The RAADS-R test does not ask about camouflaging, so adults who have masked successfully for years can score lower than their experience would suggest. This affects women disproportionately.
What is the difference between the RAADS-R and the RAADS-14?
The RAADS-14 Screen is a 14-item short form derived from the 80-item RAADS-R test, developed for use in busy psychiatric clinics. Its threshold is 14, with sensitivity of 97 per cent but specificity of only 46 to 64 per cent.
Is the RAADS-R test better than the AQ?
Neither clearly outperforms the other. The RAADS-R test is longer and takes a lifetime view; the AQ is shorter and asks about the present. Studies comparing them in clinical settings found both performed in the poor to fair range. Neither asks about masking.
Can teenagers take the RAADS-R test?
It was validated for adults aged 18 and over. For teenagers, parent-report instruments such as the SRS-2 or the adolescent version of the AQ are more appropriate.
Where can I take the RAADS-R test?
Several non-commercial sites host it, and clinicians can administer it directly. Because the licence prohibits commercial use, be cautious with sites that run advertising, and check whether reverse-scored items are handled correctly.
Should I bring my RAADS-R score to my doctor?
Bring the examples rather than the number. No clinician treats a self-reported score as evidence, but concrete examples from your own life in the areas that scored highest are exactly what an assessment needs.
Research and sources
Everything on this page comes from published research. The papers below are grouped so you can see the shape of the evidence: the original validation first, then the independent studies that tested it in real clinics, then the short forms, translations and comparison instruments. Where a paper is open access the full text is free to read.
The original RAADS-R paper
Ritvo RA, Ritvo ER, Guthrie D, Ritvo MJ, Hufnagel DH, McMahon W, et al. (2011). The Ritvo Autism Asperger Diagnostic Scale-Revised (RAADS-R): A Scale to Assist the Diagnosis of Autism Spectrum Disorder in Adults: An International Validation Study. Journal of Autism and Developmental Disorders, 41(8), 1076-1089.
The paper that introduced the instrument. 201 autistic adults were compared against 578 controls, producing the widely quoted figures of 97 per cent sensitivity and 100 per cent specificity at a cutoff of 65, with test-retest reliability of 0.987.
DOI: 10.1007/s10803-010-1133-5 ·
Find on PubMed
Studies that tested it independently
Jones SL, Johnson M, Alty B, Adamou M (2021). The Effectiveness of RAADS-R as a Screening Tool for Adult ASD Populations. Autism Research and Treatment.
The single most important paper if you want to understand what the score is actually worth. Fifty adults on a UK specialist service waiting list completed the questionnaire. Sensitivity was 100 per cent but specificity was 3.03 per cent, 98 per cent scored above the threshold while only 34 per cent were diagnosed, and the area under the curve was 0.45. The authors concluded the instrument lacks predictive validity for this purpose. Open access.
Full text on PubMed Central
Wigham S, Rodgers J, Berney T, Le Couteur A, Ingham B, Parr JR (2019). Psychometric properties of questionnaires and diagnostic measures for autism spectrum disorders in adults: A systematic review. Autism, 23(2), 287-305.
A systematic review covering the adult screening instruments side by side. Useful for seeing how the RAADS-R compares with the AQ and others rather than reading about either in isolation.
DOI: 10.1177/1362361317748245 ·
Find on PubMed
Conner CM, et al. Comparing autism screening measures in adults referred for specialist assessment.
Ninety-three adults assessed at a specialist service. Those who received a diagnosis scored higher on the clinician-administered ADOS, but no difference appeared between diagnosed and undiagnosed groups on either the AQ or the RAADS-R, and all three measures showed sensitivity and specificity in the poor to fair range.
Find on PubMed
Analyses of factor structure (Sturm et al., 2024 and related work).
Several recent analyses find the instrument behaves statistically as a single dimension rather than the four independent subscales its structure implies. The subscales remain useful descriptively, but they are not measuring four separate things.
Find on PubMed
Short forms and translations
Eriksson JM, Andersen LM, Bejerot S (2013). RAADS-14 Screen: validity of a screening tool for autism spectrum disorder in an adult psychiatric population. Molecular Autism.
How the 14-item short form was built from the original 80 items, and why a shorter instrument was needed for busy psychiatric services. Reports sensitivity of 97 per cent and specificity of 46 to 64 per cent at a cutoff of 14. Open access.
Full text on PubMed Central
German-language validation (RADS-R, 2024). A novel screening instrument for the assessment of autism in German language: validation of the German version of the RAADS-R.
Notable because participants completed it without a clinician present, unlike the original study. Reported sensitivity of 92.5 per cent and specificity of 93.6 per cent, sitting between the original figures and the poorest independent results.
Full text on PubMed Central
The instruments it is usually compared with
Baron-Cohen S, Wheelwright S, Skinner R, Martin J, Clubley E (2001). The Autism-Spectrum Quotient (AQ). Journal of Autism and Developmental Disorders, 31(1), 5-17.
The original AQ paper, the other major adult self-report measure.
DOI: 10.1023/A:1005653411471 ·
Full text (PDF)
Allison C, Auyeung B, Baron-Cohen S (2012). Toward brief “Red Flags” for autism screening: the Short Autism Spectrum Quotient. Journal of the American Academy of Child and Adolescent Psychiatry, 51(2), 202-212.
The paper behind the AQ-10, the ten-item screen used in UK clinical guidance.
DOI: 10.1016/j.jaac.2011.11.003 ·
Find on PubMed
Hull L, Mandy W, Lai M-C, et al. (2019). Development and Validation of the Camouflaging Autistic Traits Questionnaire (CAT-Q). Journal of Autism and Developmental Disorders.
Covers the gap left by both the RAADS-R and the AQ, neither of which asks about masking.
DOI: 10.1007/s10803-018-3792-6 ·
Find on PubMed
Loomes R, Hull L, Mandy W (2017). What Is the Male-to-Female Ratio in Autism Spectrum Disorder? A Systematic Review and Meta-Analysis. Journal of the American Academy of Child and Adolescent Psychiatry, 56(6), 466-474.
Relevant context for why instruments that do not ask about masking under-detect women.
DOI: 10.1016/j.jaac.2017.03.013 ·
Find on PubMed
General reference
Well-sourced starting points if a term on this page is unfamiliar.
- Autism on Wikipedia, covering characteristics, history and current understanding.
- Autism spectrum, including how the diagnostic categories have changed.
- Autism-Spectrum Quotient, the AQ, its versions and documented criticisms of specific items.
- Asperger syndrome, the former diagnosis that gives the RAADS-R its name, and why it was merged into autism spectrum disorder.
- Autistic masking, camouflaging and what it costs.
- Sensitivity and specificity, worth reading if the figures on this page were unfamiliar.
- DSM, the manual whose criteria underpin assessment in much of the world.