Where the PSQI comes from and what it captures
Buysse and colleagues published the Pittsburgh Sleep Quality Index in 1989 at the University of Pittsburgh, and it has since become one of the most cited self-report tools in sleep medicine. The original instrument asks about the past month and folds nineteen self-rated items into seven components: how you would rate your sleep overall, how long it takes to drift off, how many hours you actually sleep, how efficient that time in bed is, how often specific disturbances wake you, whether you lean on sleep medication, and how much your days suffer as a result. This page is a shorter PSQI-style screener rather than the validated questionnaire. It borrows the same 0-to-3 way of scoring the patterns most people can recall without a diary, but it will not stand in for the full instrument or for a clinician's read.
Think of the number it returns as a conversation starter, not a verdict. A low total lines up with the kind of answers people give when sleep is going reasonably well. A higher total is worth noticing, particularly when the same pattern has held for several weeks, when it bleeds into your work or your commute, or when it comes with witnessed breathing pauses, loud snoring, chest discomfort, or a growing reliance on pills to fall asleep.
Answering the six questions for a typical month
Each question asks about your last four weeks or so, not about a single rough night after a red-eye flight or a sick child. Pick the frequency that matches your ordinary week, then type in the hours you actually sleep rather than the hours you spend lying in bed hoping to. The screener adds up the five frequency answers and converts your sleep hours into a sixth component. Zero is always the healthiest response; a 3 flags the most frequent or most severe version of that difficulty.
When the result appears, read the total and the per-component list side by side, because two people can land on the same number for very different reasons. Six hours of sleep with no other complaints points toward a duration problem; a full eight hours paired with morning grogginess and reported gasping points somewhere else entirely. If your total is high, the most useful thing to do is carry the component breakdown to a clinician instead of trying to diagnose yourself from it.
How the components add up to a score
The full PSQI global score is simply the sum of its seven component scores:
Each component ranges from 0 to 3, so the full questionnaire ranges from 0 to 21. This page collects six answers and computes the seventh component from your bed and wake times, so the total runs the full 0 to 21 range. The sleep-duration component adds 0 points for 7 or more hours, 1 point for 6 to under 7 hours, 2 points for 5 to under 6 hours, and 3 points for under 5 hours. The other five questions add their selected 0 to 3 values directly.
Walking one honest month through the screener
Picture a month of shift work. You call your overall sleep fairly bad (2), you fight to fall asleep once or twice a week (2), you average about 5.5 hours a night, which lands in the 5-to-6-hour band (2), you nod off during the day less than once a week (1), you take nothing to help you sleep (0), and you notice no breathing trouble (0). Those six answers total 7. Now add the seventh component: with a bedtime of 11 pm and a 7 am alarm you are in bed for 8 hours but asleep for 5.5, an efficiency of 68.8 %, which scores 2. The global score is therefore 9 out of 21.
Nine sits well above the cutoff of 5, and the breakdown says why: quality, latency, duration and efficiency each contribute 2 while medication and breathing contribute nothing. That is a coherent picture of not enough sleep in too much time in bed, rather than of sleep being broken up by something. The practical reading is to look hard at when the shift rotation lets you go to bed, how late the caffeine runs, and whether shortening the time in bed would actually raise efficiency — and to raise it with a clinician if the pattern holds once the rotation eases.
What each score band is telling you
Simplified PSQI-style score interpretation
| Score |
Category |
How to read it |
| 0 to 5 | Lower concern | Your answers are broadly consistent with better recent sleep quality. |
| 6 to 10 | Watch and adjust | Sleep quality concerns are present; review routines and track the trend. |
| 11 to 18 | Higher concern | Consider professional follow-up, especially if symptoms persist or affect safety. |
The seven components and what each one is asking
The published instrument scores seven areas, each from 0 to 3, and this page now covers all seven. Subjective
quality is your own verdict on how well you slept. Latency asks how often it takes more than half
an hour to fall asleep. Duration converts hours of actual sleep into points. Efficiency
compares time asleep with time in bed. Disturbance covers the things that wake you or make sleep
uncomfortable. Medication records how often you take something to sleep. Daytime dysfunction
asks what the nights are costing your waking hours.
Six of those come straight from the questions on this page. The seventh, efficiency, is computed rather than asked: your
bedtime and getting-up time give the hours in bed, and your reported sleep hours give the numerator, so
The instrument scores that at 0 above 85 %, 1 from 75 to 84 %, 2 from 65 to 74 %, and 3 below 65 %.
Because all seven components are present, the total runs from 0 to 21 exactly as the published version does, which is what
makes the familiar cutoff meaningful rather than approximate.
Reading the total against the published cutoff
Buysse and colleagues set the threshold at a global score greater than 5. In their validation sample that boundary
separated poor sleepers from good sleepers with roughly 89.6 % sensitivity and 86.5 % specificity — strong
for a questionnaire, and still only a screening result. Crossing it means your pattern of answers resembles the poor-sleeper
group in that study. It does not identify which sleep disorder, if any, is responsible, and it cannot rule one out either: a
low score with loud snoring and witnessed pauses in breathing still deserves attention.
Two people can reach the same total by entirely different routes, which is why the component chart matters more than the
number above it. A total of 8 made up mostly of duration and efficiency points describes someone not allowing enough time for
sleep, and the fix is usually a scheduling one. The same 8 made of disturbance and daytime points describes sleep that is
being broken up rather than cut short, which points somewhere else entirely. The bar chart shows at a glance which
components are carrying the total, and the result line names the largest single contributor.
Sleep efficiency is the component people most often overlook
Lying awake in bed is scored, not ignored. Someone in bed for nine hours who sleeps six has an efficiency of about
67 %, which scores 2 even though nine hours in bed sounds generous. The common instinct when sleep is poor is to spend
longer in bed, and that reliably drives efficiency down. The relationship also gives a useful consistency check: reporting
more hours asleep than hours in bed is impossible, and the calculator rejects that combination rather than scoring it.
Because bedtime and wake time wrap across midnight, an 11 pm to 7 am pattern is correctly read as eight hours in
bed rather than as a negative span. If you nap, the instrument is about the main sleep period, so leave naps out of both the
hours and the times, and note them separately when you discuss the result with someone.
Sources. The component definitions, the scoring bands and the cutoff are all taken from the original instrument and its validation.
- The instrument, its seven components, the 0 to 21 range and the greater-than-5 cutoff with its reported sensitivity and specificity: Buysse DJ, Reynolds CF, Monk TH, Berman SR, Kupfer DJ. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Research 28(2):193-213, 1989.
- General sleep-duration guidance used to interpret the duration component: US Centers for Disease Control and Prevention, About Sleep.
- When self-reported sleep problems warrant instrumented assessment rather than a questionnaire: US National Heart, Lung, and Blood Institute, Sleep Apnea.
Questions about scoring and interpreting the PSQI
What does a PSQI score above 5 actually mean?
In the original 1989 validation study, a global score above 5 separated people with poor sleep quality from good sleepers with about 89.6 percent sensitivity and 86.5 percent specificity. That is a screening threshold, not a diagnosis. It tells you your answers resemble those of the poor-sleeper group in that study, which is a reason to look further, not a finding about any particular sleep disorder.
How is sleep efficiency calculated here?
It is the hours you actually sleep divided by the hours you spend in bed, expressed as a percentage. The bedtime and getting-up time fields give the time in bed, wrapping correctly past midnight, and your reported sleep hours give the numerator. The PSQI scores efficiency above 85 percent as 0, 75 to 84 percent as 1, 65 to 74 percent as 2, and below 65 percent as 3. Sleep time greater than time in bed is rejected as an inconsistency rather than scored.
Why does this page score seven components when I only answer six questions?
Because the seventh is computed rather than asked. Sleep efficiency comes out of the bed time, wake time and sleep hours you already provide, so no extra question is needed. That brings the range to 0 to 21, matching the published instrument, which is why the standard cutoff above 5 can be quoted meaningfully here. The full questionnaire gathers more detail within several components than six questions can.
Can two people with the same score have very different sleep problems?
Yes, and that is why the component breakdown matters more than the total. A score of 8 driven almost entirely by short duration and low efficiency suggests too little time allotted to sleep. The same 8 driven by disturbance and daytime sleepiness with normal duration suggests the sleep is being fragmented rather than shortened. The two point to completely different next steps, and the total alone cannot distinguish them.
Should I use my weekday or weekend sleep pattern?
The PSQI asks about a typical month, so use what is ordinary across the whole period rather than either extreme. If weekdays and weekends differ sharply, that difference is itself worth noting: a large weekend catch-up usually indicates accumulated weekday sleep debt. Scoring the two patterns separately and comparing them is more informative than averaging them into a single figure that describes neither.
Is this a substitute for a sleep study?
No. Everything here rests on your own recollection, and self-report cannot detect what happens while you are asleep. Obstructive sleep apnoea, periodic limb movements and the architecture of sleep stages are invisible to any questionnaire and require instrumented assessment. Persistent high scores, loud snoring with witnessed pauses in breathing, or dangerous daytime sleepiness are reasons to speak to a clinician regardless of what this page reports.
Limitations: what a self-report screener cannot see
Everything here rests on your own recollection, so the usual caveats about memory apply: most people underestimate how long they lie awake and overestimate total sleep. The screener never measures blood oxygen, apneas, periodic limb movements, circadian misalignment, drug interactions, pain, anxiety, or depression, and it flattens the official scoring, especially the sleep-efficiency math and the long list of nighttime disturbances the real PSQI asks about. Read the number as a rough screen, not a diagnosis or a treatment plan.
Seek qualified medical advice if high scores persist, if you regularly feel dangerously sleepy, if another person notices gasping or breathing pauses during sleep, if you fall asleep while driving, or if you are changing sedatives, stimulants, alcohol use, or other medications. Official questionnaires, sleep diaries, actigraphy, or a sleep study may be needed for a complete evaluation.