Adult ADHD self-screening tests

ASRS v1.1 — what it does, what it doesn’t

The Adult ADHD Self-Report Scale (ASRS) at its best catches two-thirds of actual adult ADHD cases. High specificity, moderate sensitivity — meaning a positive is hard to fake but a negative is not a clean ruling-out. It was built by Kessler, Adler and colleagues for the World Health Organization (WHO) World Mental Health Survey, in two related forms: a 6-item Screener (Part A) and an 18-item Symptom Checklist (Parts A + B). The screener is scored on a 5-point frequency scale with shaded boxes indicating positive responses; four or more positive items is screen positive.

The psychometrics are worth reading in fine print. Kessler 2005¹ — 6-item screener: sensitivity 68.7%, specificity 99.5% against clinician diagnosis in the National Comorbidity Survey Replication (NCS-R) clinical reappraisal subsample. High specificity means few false positives at the general population base rate. Sensitivity around 69% means roughly one in three actual adult ADHD cases is missed by the screen. Ustun 2017² re-derived the 6-item screener for DSM-5 with reported sensitivity 91.4% and specificity 96.0% — but in the same development sample. External validations in unselected adult populations report wider performance ranges. Both papers are usually administered under the “ASRS v1.1” brand, but the items have changed between them. Worth knowing which version you took.

Where it fits in the pathway: at the should I seek evaluation step. A self-administered first filter. The instrument is free, copyrighted-but-freely-distributable, and hosted as a PDF by the Harvard NCS team for WHO¹¹. The ASRS does not establish the age-of-onset criterion (before 12), the cross-setting criterion, the impairment criterion, or the differential — those need a clinician.

DIVA-5 — clinician interview, not a self-test

DIVA-5 (Diagnostic Interview for ADHD in adults)³ was developed by J.J. Sandra Kooij and colleagues in the Netherlands; the DSM-5 update of the original DIVA. It walks through all 18 DSM-5 ADHD symptoms with 4–6 concrete real-world examples per symptom for both current and childhood (before age 12) presentations, plus impairment across five life domains — work/education, relationships and family, social functioning, free time, self-image. Free for clinical use; research use requires a paid license.

There is a patient-facing self-completion version. It is intended as appointment preparation— to be brought to and reviewed with a clinician — not as a standalone diagnostic. Some adults complete it alone, conclude they have ADHD on the basis of the count, and stop there. That’s an inversion of how the instrument was designed. The interview itself is the diagnostic act; the patient version is the homework before the interview. The Updated European Consensus Statement on Adult ADHD (Kooij et al. 2019)⁴ sets out how DIVA fits in the wider European diagnostic protocol.

ADHD-RS, CAARS, BAARS-IV — when each fits

ADHD-RS-IV / ADHD-RS-5⁵ (DuPaul et al.). Rating scales matched 1:1 to DSM symptom criteria. Originally developed for children; the adult version uses the same item structure rephrased for adult contexts. Self-report or informant-report. Primary use: symptom-severity quantification at baseline and treatment-response tracking over time — not initial diagnosis on its own.

CAARS-2 (Conners’ Adult ADHD Rating Scales, 2nd ed., 2022)⁶. The most-used commercial instrument in US clinical assessment. Multimodal: self-report and observer-report (partner, parent, clinician), in long (66 items), short (26 items), and screening (12 items) versions. Roughly $200–400 for a starter kit, per-use scoring fees apply. Used clinically because the multi-rater structure produces a defensible report; less used in screening because it’s not free. CAARS-2 added updated norms and DSM-5-TR alignment.

BAARS-IV (Barkley, 2011)⁷. Covers current symptoms, retrospective childhood symptoms, and executive function via self-report and other-report. Widely used in research; modest cost (~$100 kit); normed on US adults. The retrospective childhood-symptom subscale is often used to anchor the “before age 12” DSM criterion when no collateral informant is available.

WURS (Wender Utah Rating Scale, Ward 1993)⁸. 61-item retrospective childhood-symptom scale, used to anchor the pre-12 criterion when no informant exists. Reasonable psychometrics; the scale’s age and the construct shift from DSM-III-R complicate interpretation, but it’s still the best-known retrospective-childhood instrument.

Screening positive ≠ meeting criteria

The DSM-5-TR adult ADHD diagnosis requires five or more of nine inattentive symptoms and/or five or more of nine hyperactive-impulsive symptoms (the threshold drops from six to five at age 17), several present before age 12, in two or more settings, with clear functional impairment, not better explained by another disorder. (DSM-5-TR, APA, 2022.)⁹ A self-screen establishes only the current symptom count. It does not establish age of onset, the cross-setting criterion, the impairment criterion, or the differential against other conditions that produce similar symptom profiles.

Translating that into plain advice: a positive ASRS means worth a clinician conversation. It is not, by itself, a diagnosis.

The comorbidity false-positive problem

Roughly 60–80% of adults with ADHD have at least one comorbid disorder (Kessler 2006 NCS-R¹⁰). The converse is also true: adults with depression, generalized anxiety, bipolar II, autism spectrum, PTSD, or chronic sleep deprivation routinely score positive on the ASRS because their symptom profiles overlap with the inattentive criteria.

The article’s honest implication: a positive ASRS read in isolation can’t differentiate ADHD from a treatable mimic. A good clinical evaluation pairs the ASRS with structured screening for depression (PHQ-9), anxiety (GAD-7), bipolar (MDQ), substance use (AUDIT, DAST), sleep apnea (Epworth, STOP-BANG), and where indicated, autism (RAADS-R, AQ-10). A self-screen without that surrounding workup is one data point.

Validation gaps in diverse populations

Most of the original ASRS validation used the NCS-R US sample, which was predominantly white. Subsequent ASRS validation in Latino, Black US, Asian-American, and African / Asian / Latin American populations exists but is patchy and not consistently replicated. The WHO ASRS has been translated into roughly thirty languages; most cross-cultural validation studies are small and use varying gold-standard comparators. DIVA-5 has been translated into 25+ languages but external psychometric validation per language is uneven. CAARS norms are US-white-majority.

The practical caveat: instrument cut-points calibrated on the original samples don’t straightforwardly translate. Where the reader belongs to an under-validated population, the clinician interpreting their score should know the limitation. The most defensible move is the same as in the homogeneous case: treat the screener as a signal, not the diagnosis.

The ASRS vs the online quiz that uses its items

Platform “ADHD quizzes,” reader-magazine self-tests, BuzzFeed-style symptom lists, TikTok symptom checklists, intake-form symptom checkers — none of these are validated instruments and they don’t share the ASRS’s psychometric properties. Some lift ASRS items but score them without the shaded-box logic; others use new items entirely. A positive on one of these is not equivalent to a positive on the ASRS, and a positive on the ASRS is not equivalent to a diagnosis. Two layers between “internet says I might have ADHD” and the clinical conclusion.

The pathway — what to use, when

Five decision points, five instruments. The mismatch most readers run into is using a clinician-rated severity tool as a self-diagnosis or a screener as a treatment-response tracker. Each instrument has a slot.

• Should I seek evaluation? Take the ASRS v1.1 6-item screener from the WHO PDF. Positive (≥4 shaded items) means worth booking a clinician.
• Prepare for the evaluation. Take the DIVA-5 self-completion version. Bring it to the appointment. It saves the clinician fifteen minutes, demonstrates self-knowledge, and gives them a structured starting point.
• During / after the evaluation.The clinician will probably use CAARS, ADHD-RS-5, or BAARS-IV for severity quantification and a baseline for treatment-response tracking. If they don’t propose a baseline measurement, ask why.
• If no collateral childhood informant exists. The WURS or the BAARS-IV retrospective-childhood subscale is the standard substitute. Imperfect; better than nothing.
• Tracking treatment over months. ADHD-RS-5 or CAARS short-form completed monthly during titration produces actual data to bring to the next prescriber appointment.

The pathway above is the version that produces the most defensible evaluation. Most readers will only encounter the first one — and most clinical evaluations stop short of the systematic before/after measurement at the end. Knowing what the full pathway looks like is what lets you ask for the missing pieces.