Are online ADHD quizzes reliable?

Search for "ADHD test" and you will find hundreds of quizzes: some from universities and health agencies, many from apps, telehealth companies and lifestyle sites. They look similar, but they are not equally trustworthy. This page compares the main types, explains what research says about their accuracy, and gives you a quick way to judge any quiz you come across.

Three kinds of online ADHD quiz

The adult ADHD self-test on this site uses the ASRS v1.1 Screener with its original wording and scoring, as the instrument's terms require.

What research says about validated screeners

The ASRS v1.1 Screener has been studied for two decades. In its original 2005 validation, in 154 adults from a US national survey compared with blind clinical ratings, it had sensitivity of 68.7% and specificity of 99.5%. A 2007 study of 668 health plan members found strong agreement with clinician diagnoses (area under the curve 0.90). The details, and what they mean for your own score, are in what your ADHD self-test results mean.

Note the job the authors had in mind. They recommended the screener for community surveys and for clinical outreach and case-finding: finding people who should be offered an evaluation. Confirming a diagnosis was never its purpose.

Two findings matter most for people taking a quiz online.

• Accuracy depends on who is taking it. Among 1,138 adults entering substance use treatment, the ASRS v1.1 Screener had a positive predictive value of 0.26: about one in four positive screens was confirmed. For a newer DSM-5 version of the screener, specificity was 96.0% in the general population but 74.0% in a specialty clinic. People who go looking for an ADHD quiz are, by definition, people with attention complaints, and that group includes many with anxiety, depression, sleep problems or stress.
• Self-report is easy to shape. A 2007 study in Archives of Clinical Neuropsychology asked students to simulate ADHD and found that the symptoms are easily fabricated and that simulators would be indistinguishable from those with true ADHD. Most people taking a quiz are not faking anything, but the finding shows that a self-report form cannot check itself. If you take a quiz hoping for a particular answer, it is easy to give it one without meaning to.

What about full "online ADHD assessments"?

Some companies sell longer online assessments that go beyond a screener. Published evidence on them is thin. One 2025 study in the Journal of Clinical Psychiatry compared a proprietary asynchronous online assessment with a video interview by licensed clinicians in 345 self-referred adults. The study was funded by the company that makes the assessment. Overall agreement was 78.0%, with sensitivity of 80.6% and specificity of 44.0%. Agreement beyond chance was low (kappa 0.13), and about 93% of participants had ADHD on interview, so the authors noted the results may not generalize. In other words, the assessment did reasonably at confirming ADHD in a group where most people had it, and poorly at identifying the people who did not.

That is one study of one product. It is not evidence about online assessments in general, and it is a reminder to ask any company what independent evidence supports its tool.

Social media is a separate problem

Many adults first suspect ADHD from short videos. A 2022 study in the Canadian Journal of Psychiatry analyzed the 100 most-viewed TikTok videos tagged #adhd. Clinicians rated 52% as misleading, 27% as personal experience and 21% as useful. The videos averaged 2.8 million views each. The authors warned of an increased risk of overdiagnosis or misdiagnosis among viewers. Recognizing yourself in a video is a reason to look into it, not evidence either way.

Why no quiz can diagnose

The CDC says there is no single test to diagnose ADHD. UK guidance from NICE says a diagnosis should not be made solely on the basis of rating scale or observational data, and calls for a full clinical and psychosocial assessment, a developmental and psychiatric history, and observer reports. A quiz cannot:

• check whether symptoms started before age 12, which NIMH says is required for adults;
• weigh how much the symptoms impair you across settings;
• ask follow-up questions when an answer is ambiguous;
• rule out other conditions that produce the same answers.

A quick checklist for any quiz

1. Does it name the instrument and cite the research? "Based on the ASRS" is weaker than the original questions with a citation.
2. Is the scoring explained? You should be able to see how your answers became a result.
3. Does it say clearly that it is not a diagnosis?
4. What happens to your answers? Check whether they are stored, linked to your email or shared. A screening result is health information.
5. Is the result tied to a sales pitch? Be cautious if a positive result leads straight to a paid subscription or a promise of medication. The New York guide's red flags of online ADHD clinics covers what to watch for.

How to use a quiz well

• Choose a validated screener with unchanged wording.
• Answer for how you usually are, not your worst or your easiest week.
• Take it once. Retaking a quiz until the answer changes tells you about the retakes, not about you.
• Treat a positive result as a reason to book an evaluation, and a negative result as reassuring but not final.
• Bring your answers and real-life examples to the appointment. See how to talk to a doctor about adult ADHD.

When to talk to a clinician

Whatever a quiz says, the problems that sent you looking are what matter. If attention, organization or restlessness have been causing trouble for months in more than one part of your life, an evaluation is the reliable next step. See when to get evaluated for adult ADHD.

Dr. John C. Shershow, M.D. is a psychiatrist who focuses on adult ADHD and sees adults located in New York or New Jersey. His adult ADHD evaluation reviews current symptoms, childhood, school and work history, health, sleep and substances, and other explanations. See adult ADHD testing in NYC or in New Jersey. To start, call the office or use the call-back form.