What this self-check measures
The nine questions above come from the PHQ-9 — the Patient Health Questionnaire-9 — the most used depression screen in the world.1 It doesn't ask whether you are depressed; it asks how often specific symptoms have shown up over the last two weeks: the interest that's drained out of things, the low mood, the sleep that's broken or endless, the energy that isn't there, the harsh inner verdicts.
Those nine items aren't a loose collection. Each one maps directly onto one of the nine diagnostic criteria for major depression in the DSM — the manual clinicians use. That's the PHQ-9's quiet advantage over the hundred informal "depression quizzes" online: when you answer these nine questions, you're covering exactly the ground a doctor would cover in a diagnostic interview, in about two minutes.
The two-week window matters too. Everyone has flat days — a rough stretch at work, bad news, a gray week in February. The PHQ-9 is looking for the sustained pattern: what your mood and energy have been doing across fourteen days, not on the worst one. That's why the honest answer is usually about frequency ("how often"), not intensity ("how bad").
How scoring works
Each answer is worth 0 to 3 points — from "not at all" to "nearly every day" — for a total between 0 and 27. The bands you see with your result are the same ones used in clinics:
| Score | Band | Standard reading |
|---|---|---|
| 0–4 | Minimal | Symptoms at the level of ordinary background noise |
| 5–9 | Mild | Noticeable; this is where foundations do their best work |
| 10–14 | Moderate | The level where a professional conversation is usually recommended |
| 15–19 | Moderately severe | A clear signal to seek support soon — treatment measurably helps here |
| 20–27 | Severe | Contact a healthcare professional promptly; support at this level changes outcomes |
The cutoff of 10 isn't arbitrary. In the original validation study, a score of 10 or above identified major depression with 88% sensitivity and 88% specificity1 — a rare symmetry in screening instruments. And it held up: a 2019 meta-analysis in the BMJ, unusual for pooling individual patient data from dozens of studies rather than just their summaries, confirmed that 10 remains the cutoff that best balances catching true cases against flagging false ones.2
What a score can and can't tell you
A screening score is a signal strength reading, not a verdict. It can't diagnose you — no questionnaire can — and it can't tell you why the reading is what it is.
That second limit is worth sitting with, because several PHQ-9 items have more than one possible source. Broken sleep, low energy, and poor concentration are depression symptoms — and also symptoms of six months of accumulating sleep debt, an untreated thyroid, a new baby, or grief doing its normal, brutal work. A moderate score doesn't tell you which of those is generating it. A clinician can help you sort that out, which is one more reason an elevated score points toward a conversation rather than a conclusion.
The other direction matters too: a low score doesn't overrule your own sense that something is off. Screens measure the symptoms they ask about. If your two weeks read minimal but your life doesn't feel right, that instinct deserves attention, not dismissal — the WHO-5 Well-Being Index approaches the same territory from the opposite side, asking about the presence of good things rather than the presence of symptoms, and sometimes catches what a symptom screen misses.
If your score was minimal or mild
First: believe it. People who've had hard seasons often distrust a good reading, assuming they answered wrong or caught a lucky fortnight. If your last two weeks genuinely read 0–9, your load and your capacity are roughly in balance right now.
The useful move at this level is maintenance, not intervention. The foundations — sleep, daylight, movement, food, people who matter — do more for background mood than any technique, and they're far cheaper to keep than to rebuild. In the mild range especially, these are the levers with the best track record. If one number in your life is quietly degrading, it's usually sleep; a mild score with rising sleep debt is a forecast, not a coincidence.
Then watch the trend. Moods move slowly, in both directions, and a single reading can't show you which way yours is drifting. Retake in two to four weeks. Mild-and-falling is a very different situation from mild-and-climbing, and only the second reading can tell you which one you're in.
If your score was 10 or higher
A score of 10 or more is the level where clinicians recommend a real conversation — with a doctor, a therapist, or whatever door is easiest for you to walk through. Not because the number proves something is wrong with you, but because at this symptom level, treatment and support measurably help, and waiting it out measurably tends not to.
Two practical notes from people who've been there:
- Print or share the result. The hardest part of the first appointment is often the first sentence. Walking in with "I took a PHQ-9 and scored 13" replaces five minutes of throat-clearing with a number the clinician immediately understands — it's the same instrument they use. The print button above formats your result for exactly this.
- A high score doesn't cancel your competence. Plenty of capable, high-functioning people carry depression symptoms while showing up every day, and the showing-up can hide the cost for a long time. Seeking support isn't a demotion from the ranks of people who cope. It is coping — the effective kind.
If your score landed at 15 or above, read "soon" as this week, not this quarter. And if making the appointment feels like lifting a car, borrow someone: ask one person you trust to sit with you while you book it. That's not weakness. That's logistics.
About question 9
One question on this screen works differently from the other eight, and it deserves plain language.
Question 9 asks about thoughts that you would be better off dead, or of hurting yourself. On every other item, what matters is the total. On this one, any answer above zero matters more than the total — a 1 on question 9 with an otherwise low score is more important information than a 12 without it. That's not the tool being dramatic; it's how clinicians read the instrument, and it's how this page is built: any non-zero answer on question 9 surfaces crisis resources immediately, regardless of your overall band.
If that was you — if those thoughts have been visiting, at any frequency — here is what we want you to know. Thoughts like these are a symptom, not a truth, and they are far more common than the silence around them suggests. They respond to support. People staff phone lines around the clock because these conversations work:
- US: call or text 988 (Suicide & Crisis Lifeline), or text HOME to 741741 (Crisis Text Line)
- UK & Ireland: Samaritans, 116 123
- Australia: Lifeline, 13 11 14
- Everywhere else: findahelpline.com lists free, confidential lines by country
You don't need to be in immediate danger to call. You don't need the right words prepared. "I took a depression screen and question 9 wasn't a zero" is a complete opening sentence. Calling for support in time is judgment, not surrender.
Using your history
This tool keeps your results on your device (and only there), so retakes build a private trend line. The trend is more honest than any single reading — mood scores wobble with sleep, season, and circumstance, and what you want to know is the direction underneath the wobble.
Retake every two to four weeks while you're actively working on something — a treatment, a habit change, a recovery — or quarterly as a baseline. If you also carry anxiety, the GAD-7 pairs naturally with this screen; the two conditions travel together often enough that clinicians usually check both. And the Steadiness Log tracks the other half of the picture: not how heavy the load is, but how well you're carrying it day to day.
Where the PHQ-9 comes from
The PHQ-9 was developed by Kurt Kroenke, Robert Spitzer, and Janet Williams, and validated in 2001 on some 6,000 patients across primary care and obstetrics-gynecology clinics.1 Its development was funded by an educational grant from Pfizer, which later made the instrument — along with the GAD-7 anxiety screen — freely available for use without permission, fees, or copyright restriction.3 That open license is why the same nine questions appear in health systems and research worldwide, and why we can offer them here, verbatim and free.
Our implementation adds nothing to the instrument and takes nothing away: the nine official items, the official scale, the official bands — scored in your browser, so your answers never leave your device.