Sleep Deprivation Test — The Feeling Is Not The Measurement
Search for a sleep deprivation test and you will be offered a verdict: a score, a category, a number of hours you owe. None of that is measurable from a browser tab, and the honest version of this page has to start by saying so.
What research does measure is narrower and more interesting. Under restricted sleep, responses get slower and — more tellingly — less even, with occasional long gaps where nothing happens at all. The reason that matters is the part most articles leave out: those gaps keep growing while the feeling of being fine largely stops changing. People adapt to the sensation, not to the impairment.
What you can actually measure right now
None of these returns a verdict about your sleep. Each one measures how you are responding at this moment, which is the part that can be measured honestly — and each is worth repeating on a good day and a bad one rather than reading once.
Psychomotor Vigilance Task
The task sleep researchers use, and the one where the gap between feeling fine and responding late shows up most plainly. It reports the responses you fell behind on separately from your average, because the average is where they hide.
~3 minPlay →Mackworth Clock Test
Almost nothing happens for minutes at a time, and that is the condition under which tiredness surfaces first — a task with constant demand keeps you propped up, while a quiet one lets you drift and then shows you where.
~2 minPlay →Sustained Attention to Response Task
Useful here for one specific comparison: run it when rested and again when you are not. Its timing scatter tends to widen between those two sessions even when your own sense of how well it went barely changes.
~2 minPlay →
Why there is no sleep-debt number here
To turn performance into hours of missing sleep you would need to know your own rested baseline, the time of day, how long you have been awake, your caffeine, your age, and what you were doing five minutes ago. A single session in a browser knows none of it. Any site that returns 'you are 2.3 hours short' has invented the conversion.
There is a second reason, and it is the one that would matter if the arithmetic worked. Sleep loss is not the only thing that slows responses — so does being bored, being ill, being upset, having a bad trackpad, or simply not caring much about this particular attempt. A number that lumps all of those together and labels the result 'sleep debt' is confidently wrong rather than usefully approximate.
What sleep loss actually does to attention
The measure sleep researchers reach for first is deliberately trivial: a signal appears at an unpredictable moment, you respond, that repeats for minutes. Nothing about it is hard. What changes with sleep loss is not whether you can do it but whether you keep doing it — the responses spread out, and some of them arrive very late.
Those very late ones are the interesting part. In the literature they are counted separately from the average, because an average hides them: you can keep a respectable mean while having several moments where you were, briefly, not there. Counting them separately is why the task is informative at all.
The last piece is the one worth carrying away: the responses degrade faster than the sensation does. Under sustained restriction, people rate themselves as roughly as alert as before while their responses keep slipping. That dissociation is well documented, and it is the honest reason to prefer a boring measurement over a self-assessment — not because your judgement is bad, but because that particular judgement is not built to track this.
The only comparison that means anything: you against you
Because there is no population norm here, a single score in isolation tells you almost nothing. What does carry information is the same task, done by the same person, on the same device, at comparable times of day, across several sessions. Then a change is a change rather than noise.
The version of that comparison people find most useful is deliberately unfair to themselves: run it once when you know you slept well, and once at the end of a day you know went badly. The difference between those two is yours, and it is far more informative than either number alone. Every session is saved if you are signed in, which is the entire reason the history exists.
Two practical cautions. Do not compare a morning session with a late-night one and read the gap as sleep debt — time of day moves alertness on its own, and it does so in everyone. And do not switch devices mid-comparison: a trackpad, a laptop keyboard and a phone produce different response times for reasons that have nothing to do with you.
If you are worried about your sleep
Nothing on this page is a screening instrument, and no online test can substitute for one. Persistent daytime sleepiness, loud snoring with pauses in breathing, falling asleep unintentionally during the day, or sleepiness that does not improve after several nights of adequate opportunity to sleep — those are reasons to talk to a doctor, not reasons to take another test.
That is not a formality. The tasks here are also used in clinical research, and their appearance in that context is exactly why it would be careless to let a browser score stand in for an assessment. A measurement can be real and still be the wrong tool for the question you are actually asking.
As for what helps: the honest list is short and unglamorous. More sleep, at a more regular time, is the intervention with the evidence behind it. Practising a reaction task will make you better at that task; it will not make an under-slept brain rested.
Frequently asked questions
Can this tell me whether I am sleep deprived?+
No. It can tell you how quickly and how evenly you responded during one short session, which is influenced by sleep and also by time of day, caffeine, illness, mood, your input device, and how much you cared. Separating those out is not something a single score can do.
Why do I feel fine but respond worse than last time?+
That mismatch is the best-documented finding in this area, and it is not a flaw in your self-awareness. Under restricted sleep, subjective alertness flattens out while response speed and consistency keep degrading — the sensation adapts and the performance does not. It is the main reason a boring objective task beats asking yourself how alert you feel.
How many hours of sleep do I actually need?+
That is not answerable by a test, and the honest answer is a range rather than a number: most adults need somewhere in the region of seven to nine hours, and where you fall in it is not something a browser can determine. If you want to know your own requirement, the usual approach is a stretch of days with unrestricted opportunity to sleep — not a score.
Is it safe for me to drive right now?+
This page cannot answer that and will not try. Fitness to drive is not something a website is in a position to judge, and treating a score as permission would be genuinely dangerous. If you are asking the question because you feel too tired to drive, that feeling is the more reliable signal — do not drive.
Is this a test for sleep apnea or narcolepsy?+
No, and it cannot detect either. Some of the tasks here are used as outcome measures inside clinical research on those conditions, which is a completely different thing from being able to identify them. Diagnosis involves clinical history and supervised assessment; if you suspect a sleep disorder, that is the route.