Score the DASS-21 by summing the seven items in each of the three scales, then multiplying each scale total by two. The doubled score is what you compare against the published cut-offs. Depression is normal at 0-9, anxiety at 0-7 and stress at 0-14, with mild, moderate, severe and extremely severe bands above those.
That doubling step is the most-skipped instruction in DASS-21 scoring, and missing it makes every score look about half as severe as it is. Below: the scoring mechanics, what the severity labels genuinely mean, and the two Australian rules specific to the DASS-21.
How do you score the DASS-21?
The DASS-21 is a self-report questionnaire of 21 items, split into three scales of seven items each: depression, anxiety and stress. Respondents rate how much each statement applied to them over the past week on a four-point scale, from 0 ("did not apply to me at all") to 3 ("applied to me very much or most of the time").
Scoring is arithmetic, not weighted. Add the seven items belonging to each scale, giving three raw scores between 0 and 21. Then multiply each by two. Those doubled figures, not the raw sums, are the numbers you interpret and record.
The three scales measure different things. Depression covers dysphoria, hopelessness, devaluation of life, self-deprecation, anhedonia and inertia. Anxiety covers autonomic arousal, skeletal muscle effects and situational anxiety. Stress covers chronic non-specific arousal: difficulty relaxing, nervous arousal and irritability (DASS-21 form and scoring notes, Queensland Motor Accident Insurance Commission).
Why do you multiply DASS-21 scores by two?
Because the DASS-21 is a half-length extract of the original 42-item DASS, and all the normative data sits on the 42-item metric.
When the short form was built, items were selected against three criteria: good factor loadings, coverage of every subscale within each scale, and item means chosen so that DASS-21 scale scores would land very close to exactly half the full-scale score. The authors are explicit about the consequence: "Remember when using the DASS21 to multiply the obtained scale scores by 2, so that they can be compared to the DASS normative data and to other published DASS data" (DASS FAQ, School of Psychology, UNSW).
This matters clinically, not just pedantically. An undoubled anxiety score of 9 reads as mild. Doubled, it is 18, which sits in the severe band. If your intake form reports raw sums, every anxiety result you have filed understates the person in front of you. It is worth checking once what your form stores, and whether the table you compare against assumes doubled scores.
What are the DASS-21 severity cut-offs?
These are the conventional severity bands, applied to the doubled scale scores, listed as depression / anxiety / stress:
- Normal: 0-9 / 0-7 / 0-14
- Mild: 10-13 / 8-9 / 15-18
- Moderate: 14-20 / 10-14 / 19-25
- Severe: 21-27 / 15-19 / 26-33
- Extremely severe: 28 and above / 20 and above / 34 and above
Note how narrow the anxiety bands are. The mild band is two points wide, so a one-item shift can move someone a whole category. Report the number alongside the label, and be cautious about reading a single band change between sessions as progress. The three scales are also moderately intercorrelated, with typical correlations of .5 to .7, so expect them to move together.
What do the DASS-21 severity labels actually mean?
Not what almost every summary implies. They are population-relative descriptors, not disorder severity ratings.
The authors are unusually blunt about this. The labels characterise "the full range of scores in the population", and the tip published alongside them reads: "'mild' for example means that the person is above the population mean but still well below the typical severity of people seeking help, ie it does not mean a mild level of a disorder" (DASS FAQ).
They also call the cut-offs arbitrary, and publish the labels only in the paid DASS manual (AUD$55) specifically "to avoid misinterpretation and reification of these (arbitrary) labels", which is why the tables circulating online are third-party reproductions rather than the primary source.
So a "moderate" depression result says where this person sits relative to the general population, not that they have moderate depression. Write it up that way: "DASS-21 depression 18, in the moderate band relative to population norms" is defensible. "The DASS-21 indicates moderate depression" is not.
Does a DASS-21 score give you a diagnosis?
No. The DASS was built on a dimensional rather than categorical conception of psychological disorder, on the assumption that the difference between general-population and clinical distress is one of degree. It "has no direct implications for the allocation of patients to discrete diagnostic categories postulated in classificatory systems such as the DSM and ICD".
The mapping to diagnostic categories is also less intuitive than the scale names suggest. The DASS anxiety scale corresponds most closely to the anxiety disorders with the exception of generalised anxiety disorder. It is the stress scale that corresponds quite closely to the DSM symptom criteria for GAD. A clinician screening for generalised anxiety and reading only the DASS anxiety scale is reading the wrong column.
This explains why a DASS-21 and a GAD-7 can disagree without either being wrong, and the same applies to the PHQ-9 and the K10. When building a psychology intake assessment, choose the measure you will keep using rather than administering several and reconciling them later.
The original validation work is Lovibond and Lovibond (1995), Behaviour Research and Therapy 33(3):335-343 (PubMed 7726811). The most-cited DASS-21 normative paper, Henry and Crawford (2005) in the British Journal of Clinical Psychology 44(2):227-239 (PubMed 16004657), drew its sample of 1,794 from the general adult UK population, which is worth knowing before you describe those figures as Australian norms.
Where does the DASS-21 fit under Medicare in Australia?
This is the DASS-21's strongest structural position, and it is specific to Australia.
Preparing a GP Mental Health Treatment Plan under MBS item 2715 requires an assessment process that includes "the administration of an outcome measurement tool (except if considered clinically inappropriate)". The choice of tool sits at the practitioner's clinical discretion, but the descriptor names only two examples: the Kessler Psychological Distress Scale (K10) and "DASS 21 (Depression, Anxiety and Stress)" (MBS item 2715, Department of Health).
The trap is in the review. A plan review must include "re-administration of the outcome measurement tool used in the assessment stage, except where considered clinically inappropriate". The tool is pinned at assessment. If the GP built the plan on a K10 and the treating psychologist tracks only a DASS-21 across ten sessions, the review has no comparable measure to re-administer, and nothing the psychologist collected can close that loop.
The answer is not to abandon your preferred instrument, but to find out which tool the plan was built on, collect that one for the review, and run whatever else you need alongside it. Our guide to mental health treatment plan referrals covers what else travels with the referral, and the measure should be visible in the treatment plan you write from it.
Can you put the DASS-21 on your practice website?
Almost certainly not, and this is the compliance point most practices miss when they build digital intake.
The DASS is public domain: permission is not needed, the questionnaires and scoring key may be copied without restriction, and you may charge patients for an assessment that includes it. But the authors set explicit conditions on electronic administration.
If a website or app "is intended for or open to members of the public (for example a clinical practice website), then no". If it is restricted to a defined group, such as enrolled patients, and the results are fed back to the clinician rather than the respondent, then yes. Computerised administration comes with a further instruction: "it is important that computed scores are not made available to respondents, and in particular that you do not attempt to provide automated interpretation. This can be misleading and potentially dangerous" (DASS FAQ).
So an open "check your mental health" widget on a practice homepage that returns a severity label to the visitor breaches both conditions at once. A gated intake form sent to a booked client, scored into the clinician's file and not displayed back to the client, meets them. PractaLuma is AI-native practice management software for Australian mental-health practices, and its standardised assessments are built to that gated pattern: the DASS-21 goes to an identified client and the scored result lands in the clinician's record.
What does the DASS-21 not tell you?
Four things worth stating in a report before someone else asks.
It has no lie scale. Like other transparent self-report symptom measures it can be faked good or bad, so if you expect biased responding you need a separate instrument designed to detect it.
Its lower age limit is 14, assuming typical language skills. Below that, the youth version (DASS-Y) is the appropriate instrument.
One missing item on a scale can be handled by averaging the remaining items for that scale. Too many compromises validity, and in clinical work the reason for the gap should be explored rather than imputed away.
The total score is not a raw sum. Combining the scales is sensible in principle, but the authors' method is to convert each scale to a Z score against the normative data and average those, which keeps the composite connected to the norms.
Frequently asked questions
Is the DASS-21 free to use? Yes. The questionnaire is in the public domain and may be downloaded from the DASS website and copied without restriction, and no permission is required. The severity cut-offs and detailed norms are published only in the DASS manual, which costs AUD$55. You may charge patients for an assessment that includes the DASS, but you may not sell the scales themselves.
What is the maximum DASS-21 score? Each scale has seven items scored 0-3, giving a raw maximum of 21 per scale, or 42 per scale once doubled. There is no single meaningful overall maximum, because the three scales are interpreted separately against their own cut-offs rather than summed.
How often can you re-administer the DASS-21? The standard wording asks about the past week, so weekly is the shortest interval that avoids overlapping reference periods. You may change the timeframe, but the authors note the resulting data is then no longer strictly comparable to the normative data, and the change should be disclosed when the scores are reported.
Who can interpret a DASS-21? No special skills are needed to administer it, since it is self-report. Interpretation is different: it should be carried out by someone with appropriate training in psychological science, including emotion, psychopathology and assessment. Where the respondent has sought professional help or is showing high distress, the authors specify an appropriately qualified health professional such as a clinical psychologist. The same care applies when scores enter a written assessment report.
This article is general information for practitioners about instrument scoring and Australian funding rules. It is not clinical advice, and it does not replace the DASS manual, your professional judgement, or current MBS advice. Confirm MBS requirements against MBS Online before relying on them.