The ASRS v1.1 Screener (Part A) has two scoring methods. The original: count ticks in the darkly shaded boxes, and four or more of six screens positive. The 2024 Harvard update: score each item 0 to 4 and sum to 0-24, with 14 or more positive. Part B has no total score. Neither method diagnoses ADHD.
Below: which boxes count as shaded, why the authors changed the method in 2024, what Part B is for, and why "ASRS" on a referral may not mean the ADHD scale.
What is the ASRS v1.1?
The Adult ADHD Self-Report Scale version 1.1 was developed by a World Health Organization work group alongside the WMH-CIDI diagnostic interview. The full Symptom Checklist has 18 questions, one for each DSM-IV Criterion A symptom of ADHD, each asking how often the person has felt or behaved that way over the past six months. The response options are Never, Rarely, Sometimes, Often and Very often. It is intended for adults aged 18 and over.
Six of the 18 questions form the Screener, which is also Part A of the full checklist. They were chosen by stepwise logistic regression because they best predicted a blind clinical diagnosis. The remaining 12 make up Part B.
The validation paper (Kessler et al., 2005, Psychological Medicine) compared ASRS responses against blind clinical ratings in 154 adults from the US National Comorbidity Survey Replication. The six-question screener did better than all 18 questions. Its sensitivity was 68.7% (56.3% for the full 18), specificity 99.5% (98.3%), and kappa 0.76 (0.58). That result is why Part A, not the full checklist, is the screening tool.
How do you score Part A with the shaded boxes?
The original rule is dichotomous: each question scores 1 if the answer falls in a darkly shaded box, and 0 otherwise. The shading threshold is not the same for every question, and that is the most common scoring mistake.
- Questions 1, 2 and 3 (finishing the final details, getting organised, remembering appointments) are shaded from Sometimes upwards. Sometimes, Often or Very often each score 1.
- Questions 4, 5 and 6 (delaying tasks that need thought, fidgeting, feeling driven by a motor) are shaded only from Often upwards. Sometimes scores 0.
Add up the six items for a total from 0 to 6. On the official form, four or more means the symptoms may be consistent with adult ADHD and an evaluation is worth discussing (Harvard NCS ASRS screener).
A worked example shows why the threshold matters. A client answers Sometimes to every question. Under the correct key, questions 1 to 3 score 1 each and questions 4 to 6 score 0, so the total is 3: a negative screen. If someone applies a single "Sometimes or above" rule to all six items, the same answers give 6 and a clearly positive screen. Paper forms make this obvious. Rebuilt forms and spreadsheets often lose the shading.
What changed with the 2024 ASRS scoring update?
On 28 February 2024 the Harvard team posted a scoring update. It says the 0-6 rule gave high and inconsistent prevalence estimates. It was built to be quick in primary care, but a different method held up better.
That alternative comes from a 2007 validation in 668 US health plan members. In that study a four-category version of the screener had an area under the ROC curve of 0.90 against clinician diagnoses. Internal consistency was 0.63 to 0.72 and test-retest reliability 0.58 to 0.77.
The 0-24 method works like this:
- Score every Part A item on the same scale: Never 0, Rarely 1, Sometimes 2, Often 3, Very often 4.
- Sum the six items for a total from 0 to 24.
- A total of 14 or more screens positive.
The update also groups totals into four bands:
- 0-9: low negative
- 10-13: high negative
- 14-17: low positive
- 18-24: high positive
The bands are more useful than a yes/no answer. A total of 13 is technically negative, but it sits in the band right below the cut-off. That is worth noting in the file and following up at interview, not treating as "no ADHD".
Which method should a clinic use?
The update describes the 0-24 method as more robust and better for research and prevalence work. The shaded-box method is still printed on the official form and is what most clients will have seen. A sensible approach is to record which method was used, report the 0-24 total and band where you have item-level answers, and never compare scores across methods. A "4" on the 0-6 scale and a "14" on the 0-24 scale are not the same number in different units. They come from different scoring rules.
How is Part B of the ASRS scored?
It isn't, at least not as a total. The Symptom Checklist instructions state that no total score or diagnostic likelihood is used for the 12 Part B questions. They give extra clues and act as probes for the clinical interview, and clinicians are told to pay particular attention to answers in the dark shaded boxes.
Part B is shaded too, again with uneven thresholds:
- Questions 9, 12, 16 and 18 (concentrating when spoken to directly, leaving your seat, finishing other people's sentences, interrupting) are shaded from Sometimes upwards.
- Questions 7, 8, 10, 11, 13, 14, 15 and 17 are shaded only from Often upwards.
Some tools add all 18 items into a single total. The 2005 paper did test this, and summing the 18 shaded-box scores was the best way to score the full checklist, but it was still less accurate than the six-item screener. The checklist instructions give no cut-off for an 18-item total. If a report reads one against a threshold, ask where that threshold came from.
What is the ASRS-5, and can you score v1.1 for DSM-5?
The ASRS-5 (Ustun et al., 2017, JAMA Psychiatry) is a separate six-question screener built for DSM-5 criteria. It uses integer item weights chosen by a machine-learning model. It is not the v1.1 screener with a new cut-off, and its scoring rules do not carry over.
In the general-population data, weighted to an 8.2% prevalence, it reached 91.4% sensitivity and 96.0% specificity at its threshold, with a positive predictive value of 67.3%. In a specialist adult ADHD clinic sample, where 57.7% of patients met criteria, sensitivity stayed at 91.9% but specificity fell to 74.0%. That is a useful reminder that screening accuracy depends on who you are screening. A clinic that only sees people already worried about ADHD will get more false positives than the headline figures suggest.
A note for anyone who rescored old v1.1 data for DSM-5: in a December 2023 memo, Kessler withdrew the 2018 guidance on DSM-5 scoring of the DSM-IV screener items. The 2018 memo had reversed the scoring rules for two items, and the method overfitted and substantially overestimated prevalence. The team now recommends the original DSM-IV scoring rules for data collected with the v1.1 items.
Is the ASRS the same as the Autism Spectrum Rating Scales?
No, and the shared acronym causes real confusion in referral letters and assessment reports. The Autism Spectrum Rating Scales (Goldstein and Naglieri, published by MHS) is a commercial, norm-referenced, multi-informant measure of autism-related behaviour for young people aged 2 to 18. It reports T-scores and percentiles, not a shaded-box count or a 0-24 total.
If a document says "ASRS" without a version number or a scale name, check before interpreting it. "ASRS v1.1" or "Adult ADHD Self-Report Scale" means the ADHD screener. T-scores across Social/Communication or Unusual Behaviours scales means the autism measure. When you write your own reports, give the full name the first time (our assessment report writing guide covers this).
What does a positive ASRS screen mean in practice?
It means a full assessment is warranted, not that the person has ADHD. The Harvard background memo is explicit that the screener "is only a screening scale, not a diagnostic test". People who screen positive need a diagnostic assessment by a trained clinician.
The same memo makes two further points that matter outside the US. The calibration rules come from US samples and may not hold in other countries or populations. The translations were produced with the standard WHO back-translation protocol, but no cultural validations had been done. Australian clinicians should treat the cut-offs as reasonable starting points, not local norms.
A complete adult ADHD assessment also needs things no self-report screener captures: evidence that symptoms began in childhood, impairment across more than one setting, and other explanations ruled out. Anxiety, depression, sleep problems and substance use can all produce similar symptoms. The Australian Evidence-Based Clinical Practice Guideline for ADHD (AADPA, October 2022, approved by the NHMRC) is the local reference for how identification and diagnosis should fit together. For setting treatment targets after diagnosis, see our ADHD treatment plan goals and objectives.
Can you use the ASRS to track progress?
The 0-24 total has a much finer range than the 0-6 count, so it can show broad symptom trends if you give it at consistent intervals. Keep in mind that it asks about the past six months, so repeating it monthly means the recall windows overlap heavily. It was validated for screening, not as an outcome measure. If you use it this way, record the scoring method and the date each time. Pair it with a functional measure, as you would with the DASS-21.
Can you copy the ASRS into your own forms or software?
The Harvard NCS site says the six-question screener is free to use without formal permission. Users are asked to cite Kessler et al. (2005), include the copyright notice, and not alter the instrument. That means keeping all response options, the original scoring algorithm, and the two levels of shading. The 18-question Symptom Checklist and the ASRS-5 both need a permission request through the links on that page.
PractaLuma is AI-native practice management software for Australian mental-health practices. Its standardised assessments library includes the ASRS v1.1 alongside other adult ADHD measures such as the WURS-25, so clients can complete it before the session and you can review item-level answers at intake (our intake assessment guide). See features and pricing for more.
FAQ
What is a normal ASRS score?
There is no "normal" score in the sense of population norms. On the six-item screener, three or fewer shaded-box ticks is a negative screen. On the 0-24 method, 0-9 is low negative and 10-13 is high negative.
Is a score of 4 on the ASRS high?
On the 0-6 shaded-box method, yes: four is the positive screening threshold. On the 0-24 method, a total of 4 is low negative. Always check which method produced the number.
Can a psychologist diagnose ADHD from the ASRS?
No. The ASRS is a screener. Diagnosis needs a full clinical assessment that includes developmental history, impairment across settings and other explanations ruled out.
Is there an ASRS for adolescents?
The official form says the v1.1 screener is intended for people aged 18 and over. For younger clients, choose a measure validated for their age group rather than applying the adult cut-offs.
This article is general information for clinicians and is not clinical advice. Scoring rules are taken from the instrument authors' published materials as at October 2026.