Are online ADHD tests reliable?

You search "ADHD test online" and 50 results stare back. Some are free. Others want $200 for an "instant diagnosis." Which ones actually work, and which are just taking your money?

The good ones are scientifically validated and free. Then there's everything else.

A screening is a filter, not a verdict

A screening test doesn't diagnose you. It filters. It tells you either "this is worth looking into" or "probably not ADHD," and that's the whole job.

Take the ASRS, the Adult ADHD Self-Report Scale built by the WHO and Harvard. It has 90.4% diagnostic accuracy (Brevik et al., 2020). The 6-question short version runs 68.7% sensitivity and 99.5% specificity. In plain terms, a positive result means there's a real reason to dig deeper, and a negative result means ADHD is unlikely.

ASRS, the one that's actually validated

The ASRS is the gold standard for online screening. The World Health Organization built it alongside researchers at Harvard Medical School, and a lot rides on that pedigree.

It's free and available on plenty of platforms. It's quick, 6 questions in the short version or 18 in the full one. Studies across multiple languages and populations keep confirming its accuracy. And it's been kept current, with the 2024 version switching to a Likert scale of 0-4 instead of a yes-or-no format for better precision. A 2024 Swedish study found an AUC of 0.808-0.817 in clinical populations with overlapping symptoms, which means it holds up even when comorbidities are in the picture.

The other tools worth knowing

The DIVA-5, the Diagnostic Interview for ADHD in Adults, is a semi-structured interview built on the DSM-5. It posts 100% diagnostic accuracy against the CAADID, the clinical gold standard, with 91.30% sensitivity and 93.62% specificity in the Korean version. Clinicians use it, and a self-administered version exists too.

A few more, the WURS-25, the CAARS, and the TRAQ-10, are scientifically validated as well, just harder to find online.

How to spot a scam test

A 2025 comparative study put 74 online assessments under the microscope. The mean accuracy came to 76%, but the real problem was the low positive predictive value, somewhere between 10% and 30% at actual prevalence rates.

Run from any test that promises an "instant diagnosis" for money, won't say which tool it uses, takes less than 2 minutes, or offers you medication the moment the test ends.

This isn't a hypothetical risk. Done Global, a telehealth company, was charged with fraud in 2024 after prescribing over 40 million stimulant pills off the back of "1-minute assessments." Its CEO and clinical president were arrested. Other companies, Cerebral among them, face lawsuits over rushed diagnoses and overprescription.

The self-diagnosis argument

Roughly 25% of adults suspect they have undiagnosed ADHD, and getting a formal diagnosis is genuinely hard, between months-long waitlists, prohibitive costs, and clinicians who don't grasp the adult presentation.

The science backs the seriousness of the question. Adult ADHD has well-established descriptive, predictive, and concurrent validity. The symptoms don't evaporate at 18.

So an informed self-diagnosis is a valid place to start, as long as you're using validated tools like the ASRS or DIVA, drawing on scientific sources, and treating it as a way to understand how you function rather than a license to self-medicate, especially when formal diagnosis is out of reach. The problem was never people assessing themselves carefully. The problem is medical gatekeeping that waves ADHD away because "you got good grades" or "you don't seem hyperactive."

Where online tests genuinely fall short

A 2025 study on online assessment in adults found 78% agreement with clinical evaluation. The positive predictive value was 94.9%, the negative predictive value just 15.1%. So a positive result is probably right, and a negative result you don't believe is worth a second opinion.

No online test can rule out the comorbidities that imitate ADHD, like anxiety, depression, or autism. None can weigh your full life context, and none can stand in for clinical judgment shaped by real experience with adult ADHD. A screening only catches signals. A diagnosis takes a full evaluation, your childhood history, the functional impact across several areas of life, and the work of ruling out other causes.

What to do with a positive result

1. Don't panic and don't celebrate, it's information, not a sentence
2. Write things down, with concrete examples of how the symptoms show up day to day
3. Seek a professional evaluation from a psychiatrist or psychologist who specializes in adult ADHD
4. Bring your completed ASRS, it helps the clinician rather than offending them
5. If access is genuinely impossible, use what you've learned to understand your own functioning and to find evidence-based strategies, not unsupervised medication

Free, validated tests

You can take the ASRS for free at Mental Health America, ADDA (the Attention Deficit Disorder Association), Psychology Tools, and NeuroDirect (which runs the ASRS-5). They all use the same WHO-validated tool, so pick whichever you like. It takes 3-10 minutes.

The bottom line

Online ADHD tests aren't all garbage. The ASRS is scientifically validated, free, and genuinely useful as a first step. It doesn't replace a professional diagnosis. But when the medical system fails you, and it does so constantly, a validated screening paired with informed self-knowledge beats sitting in the dark with no answers at all.

The tests aren't the problem. The problem is a system that makes diagnosis hard to reach and then blames people for figuring it out on their own.