Rating scales
Score the Montgomery-Åsberg Depression Rating Scale: ten items on 0–6, a 0–60 total, the Snaith severity bands, the ≤10 remission threshold, and an explicit flag on item 10.
Free · No sign-up · Runs entirely in your browser
Rate each item on the 0–6 MADRS anchors: 0 = symptom absent, 6 = most severe. Defined descriptions sit at 0, 2, 4 and 6; the odd values (1, 3, 5) mark intermediate states.
Total score
0 / 60
Sum of all ten items
Severity band
Normal / symptom-free
Snaith et al., 1986
Item 10 (suicidal thoughts)
None
Independent of the total
Enter the pre-treatment total to see change from baseline and the 50% response criterion.
Remission (≤10)
Yes
Where the person is now
Plain text: every item with its code, the total, the band, the remission status and the item-10 status.
Items follow Montgomery & Åsberg (1979); severity bands follow Snaith et al. (1986) and the remission target of 10 or below follows Hawley et al. (2002). This is a scoring aid for education and research support, not a diagnosis: the total measures severity, and remission rates are not comparable across studies that set the cut-point differently, so state the threshold you applied. A non-zero item 10 needs direct clinical follow-up whatever the total says; if there is immediate risk, contact local emergency services or a crisis line. Rate the items above, or load the example, to see the score update.
Reference · for the curious
The Montgomery-Åsberg Depression Rating Scale was published in 1979 with a specific design goal: to be sensitive to change during treatment. Montgomery and Åsberg selected their ten items from a larger scale precisely because those items moved most when patients improved. That origin explains both the scale's structure and why it is so often chosen as a trial endpoint.
Each of the ten items is rated 0–6 against defined anchor descriptions at the even points, with the odd numbers available for intermediate judgements. The total runs from 0 to 60. The uniform range is a practical advantage over the HAM-D's mixed 0–4 and 0–2 items: raters apply one mental model across the whole instrument rather than switching.
The bands proposed by Snaith and colleagues in 1986 are 0–6 recovered, 7–19 mild, 20–34 moderate, and 35 or above severe. Remission is conventionally a total of 10 or below, though some trials use 12; response is conventionally a reduction of at least 50% from baseline, which is why the scorer above accepts a baseline score.
Nine of the ten items are rated from what the patient reports. Only apparent sadness — the despondency communicated by posture, expression and voice — is rated purely from observation. That single observational item is what makes the MADRS a clinician-rated instrument rather than a structured questionnaire, and it is the item on which raters most often diverge, because it asks for a judgement rather than a report.
The MADRS concentrates on core mood and cognitive symptoms — sadness, tension, lassitude, inability to feel, pessimistic and suicidal thoughts — with one sleep item and one appetite item. The HAM-D, by contrast, carries three separate insomnia items, so sleep alone can contribute 6 points to its total — more than depressed mood's maximum of 4.
The practical consequence is that the MADRS is less likely to move for reasons unrelated to mood, which matters when a change score is your primary outcome. Our comparison of MADRS versus HAM-D works through the decision, and the HAM-D calculator shows the sleep-weighting issue directly.
Equipercentile linking between the two scales exists and is useful for interpreting the literature, but only percentage change is genuinely scale-invariant. Total-to-total conversions depend on the sample they were derived from and should be reported as approximations, not equivalences.
Item 10 rates suicidal thoughts, from enjoyment of life through active preparations. Like item 9 of the PHQ-9, it needs attention independent of the total: a patient can score in the mild band overall and still endorse item 10 at a level that requires immediate action. The scorer above flags it separately for that reason.
If you or someone you are assessing is at risk, contact local emergency services or a crisis line now. This page is not a substitute for clinical judgement or for care.
A self-report version covering nine of the items exists and is widely used, particularly for repeated monitoring between appointments. It is not interchangeable with the clinician-rated instrument — it omits apparent sadness, the one observational item — and scores from the two should not be pooled or compared directly. Our discussion of clinician-rated versus self-report scales covers what changes when the rater changes.
Rating from a recording rather than in the room is increasingly common in research and in centralised trial ratings, and it changes the instrument subtly: apparent sadness is available from audio and video but not from a text transcript, so a transcript-only rating of item 1 is a judgement about verbal content, and should be documented as such.
Where multiple raters are involved, agreement is the thing to measure. Because MADRS items are ordered, use weighted kappa or an ICC rather than unweighted kappa — treating a one-point disagreement as equivalent to a five-point one discards most of the information. Our guides to the MADRS items and scoring bands and scoring scales from the transcript cover the method; the inter-rater reliability calculator computes the agreement.
Montgomery and Åsberg reported high inter-rater reliability, and — in the comparison that motivated the scale's construction — better discrimination between antidepressant responders and non-responders than the Hamilton scale. Later work is consistent: Corruble and colleagues (1998) found an intraclass correlation of 0.86 for MADRS totals between raters.
Reported internal consistency is generally good, and the uniform 0–6 range is usually credited for it: raters apply one mental model across all ten items instead of switching between a 0–4 and a 0–2 scale as they do on the HAM-D. Consistency of rating is a design property here, not only a training outcome.
The MADRS is freely reproducible with citation and is hosted in the Tagaroo Scale Library.
This is a scoring aid for education and research support, not a clinical decision tool. Note in particular that item 1, apparent sadness, is an observational rating: no calculator can supply it, and a score assembled without it is incomplete. All calculation happens in your browser; nothing is transmitted or stored.
Montgomery and Åsberg (1979) reported high inter-rater reliability for the scale and, in the comparison that motivated its design, better discrimination between antidepressant responders and non-responders than the Hamilton scale. Later work supports that: Corruble and colleagues (1998) found an intraclass correlation of 0.86 for MADRS total scores between raters. Reported internal consistency is generally good, and the uniform 0–6 item range is usually credited with making the scale easier for raters to apply consistently than the HAM-D's mixed ranges.
Each of the ten items is rated 0–6 against defined anchor descriptions at the even points, giving a total from 0 to 60. The bands proposed by Snaith and colleagues are 0–6 recovered, 7–19 mild, 20–34 moderate and 35 or above severe. Nine of the ten items are reported symptoms; only apparent sadness is rated purely from observation.
The MADRS was designed to be sensitive to change during treatment, and it concentrates on core mood and cognitive symptoms rather than somatic ones — one sleep item and one appetite item, against three sleep items in the HAM-D. That makes it less likely to move for reasons unrelated to mood, which is why it is often preferred as a primary endpoint in antidepressant trials. It also has a uniform 0–6 scoring range, which raters find easier to apply consistently than the HAM-D's mixed ranges.
A total of 10 or below is the usual remission threshold, with some trials using 12. Response is conventionally a reduction of at least 50% from baseline. As with any rating scale, report the threshold you used: remission rates are not comparable across studies that define the cut-point differently.
The MADRS, published by Stuart Montgomery and Marie Åsberg in 1979, is freely reproducible with citation. A self-report version, the MADRS-S, covers nine of the items.
Written by Enrique Gutiérrez, PhD (Computer Science) — founder of Tagaroo and Associate Professor of Computer Science, working on inter-rater reliability, measurement and annotation methodology (ORCID).
How this page is checked. Scoring rules, item ranges and thresholds are transcribed from the instrument's cited primary sources and covered by automated tests that reproduce each paper's own worked examples. This page has not been reviewed by a licensed clinician, and it is a scoring aid for education and research support rather than a clinical decision tool.
Last verified: 29 July 2026. Formulas, thresholds and cited figures on this page were checked against their original sources on that date. Every calculation runs in your browser; nothing you enter is transmitted or stored.
Tagaroo lets you score the MADRS against the interview itself, linking each item rating to the passage that supports it — so ratings can be compared across raters and audited later.