The DASS-21 in one page: three 7-item scales, how to score them and why you multiply by 2, what the authors say about cut-offs, reliability and citation.
Three 7-item scales measuring negative emotional states over the past week: depression, anxiety and stress (tension). Public domain, with Spanish translations hosted on the official website and a validation in Spanish university students.
By Víctor Ciudad-Fernández, PhD · Updated: October 6, 2026 · Leer esta ficha en español · Instruments
The DASS was built with factor analysis to better separate three related negative emotional states that other questionnaires tend to blur: depression, anxiety and tension or stress, understood as persistent, non-specific arousal. It has three 14-item scales (the 42-item DASS) and a short form with 7 items per scale, the DASS-21. In P. F. Lovibond and S. H. Lovibond’s (1995) study of 717 people, the three-factor structure held in both exploratory and confirmatory factor analysis, and the scales separated the factors better than the Beck inventories.
The depression scale covers dysphoria, hopelessness, devaluation of life, self-deprecation, lack of interest, anhedonia and inertia. The anxiety scale covers autonomic arousal, skeletal muscle effects, situational anxiety and the subjective experience of fear. The stress scale covers difficulty relaxing, nervous arousal and being easily upset, irritable or impatient.
Its authors see it as a dimensional measure: they hold that the differences between the depression, anxiety and stress of the general population and those of people with a disorder are differences of degree, not kind, so they do not tie it to diagnostic categories. In general, they consider the full DASS preferable for clinical work and the DASS-21 for research.
Each item is scored from 0 to 3 according to the severity or frequency with which the person experienced that state over the past week, and each scale is the sum of its seven items (0 to 21). DASS-21 items were selected so that their score is very close to half the full DASS score, so the authors say to multiply DASS-21 scores by 2 before comparing them with their normative data or severity labels.
The instrument was built to give three separate scores. Henry and Crawford (2005) also showed that the three share a general distress factor, so a total score makes sense as an index of general distress, not as a substitute for the three scales.
The three-factor structure has been replicated in clinical and community samples: Antony et al. (1998) also found it in the DASS-21, with internal consistency and concurrent validity in the acceptable to excellent range. With 1,794 adults from the UK general population, Henry and Crawford (2005) found the best fit for a four-factor model: a general psychological distress factor plus three independent specific factors for depression, anxiety and stress. Their conclusion is that the subscales measure what they claim to, but each also taps a general dimension of distress or negative affect.
In Spanish university students, Bados, Solanas and Andrés (2005) found that the three-factor model fit the DASS-21 best, with an acceptable but not quite good enough fit; for the 42-item DASS no model reached an acceptable fit.
The authors have defined severity labels (normal, mild, moderate, severe and extremely severe) for each scale, but they publish them only in the manual and explain why: depression and anxiety vary along a continuum, so any “clinical” cut-off is arbitrary, and they do not want the labels to be reified. For research they recommend working with the scores rather than splitting samples into normal and clinical.
Out of respect for that decision, this page does not reproduce the cut-offs: if you need them, they are in the manual, and remember to apply them to DASS-21 scores multiplied by 2. One more caveat from the authors: “mild” means above the population mean but well below people seeking help, not a mild disorder.
There are several Spanish translations, all hosted on the official DASS website: Daza et al. (2002), validated with a Hispanic sample in the United States; Bados et al.’s translation of the full DASS, from which the DASS-21 is extracted; García Valenzuela’s; and IQVIA’s. The rule for any translation is that, like the original, it must stay in the public domain.
Bados, Solanas and Andrés’s (2005) validation with 365 students compares their translation with Daza et al.’s: of the 21 items, 16 are identical or very similar, 4 somewhat different and 1 very different, and they back their version with a back-translation into English. If you use a Spanish version, say which one in your methods.
License: Public domain. The official website puts it this way: no permission is needed to use it, and the questionnaires and scoring key can be downloaded and copied without restriction. The manual, with normative data and severity labels, is sold for a nominal fee.
Everything is on the DASS website at the University of New South Wales (UNSW): questionnaires, scoring, translations and the authors’ FAQ, which answers many practical questions.
Cite the manual (the source of the instrument) and the language version you use:
@book{lovibond1995dass,
author = {Lovibond, S. H. and Lovibond, P. F.},
title = {Manual for the {Depression Anxiety Stress Scales}},
edition = {2},
publisher = {Psychology Foundation},
address = {Sydney},
year = {1995}
}With seven items per scale and a general distress factor running through all three, the alpha of each subscale can tell only part of the story. If you run a confirmatory factor analysis, also report omega with the McDonald’s omega calculator, which computes it from your factor loadings.
Because the normative data and severity labels in the manual are based on the 42-item DASS, and DASS-21 items were chosen so that their score is almost exactly half. If you are not comparing with those norms or using the labels, you can work with raw scores, but say so in your methods.
According to its authors, it is suitable for screening adolescents and adults without a disorder, and they see no compelling case against using it from age 12 given adequate language proficiency, although the development samples started at 17. There is also a youth version, the DASS-Y, on the same website.
This fact sheet is an academic reference for information purposes: it does not reproduce any questionnaire item. Using the instrument in your research or practice may require citing its authors or asking for permission: see “How to get the instrument”.