Choose a test duration of 3, 5, or 10 minutes, and choose how long the waiting gaps between trials should be
The standardized combination is 10 minutes with the 2-10 second interval — that is the version used in sleep research, and the results screen tells you when you have run it. The shorter durations and the other intervals are practice variants, useful for a quick check but not directly comparable to published PVT data
The session ends once the clock reaches zero and you have answered one final trial
Press Start Test
Watch the empty panel and wait — the delay is deliberately unpredictable
A millisecond counter starts running when the trial begins
Tap anywhere in the dark test area, or press Space or Enter, as fast as you can
The counter freezes on your reaction time for one second, then the next trial begins
Keep going until the timer runs out to see your full performance breakdown
The session closes on a response rather than on the clock — when the timer reaches zero you are always given one last trial, and the test ends after you answer it, which is how the standard protocol closes a PVT
If you leave the page or open the End Test dialog, the session is held exactly where it was — a running counter resumes at the time it had reached, and a waiting gap resumes with the time it had left, so neither a slow response nor the standard interval can be altered by interrupting the test
About the Psychomotor Vigilance Task:
The PVT is one of the most widely used laboratory measures of sustained attention and alertness
It was developed by sleep researchers David Dinges and John Powell in 1985, and versions of it have since been used in NASA, military, and clinical sleep research
Its defining feature is the unpredictable wait — there is nothing to anticipate, so the task is hard to fake or pace
The wait interval is measured from your last response to the next counter, and the one second of feedback you see after each response sits inside that interval rather than on top of it
It is sensitive to sleep loss, time awake, and circadian timing
Lapses are responses slower than 500ms and are the measure most affected by sleep loss in research settings
False starts are responses made before the counter appears, or faster than 100 ms after it appears — the human visual system cannot react that quickly, so those responses are anticipation rather than reaction and are excluded from your reaction time data
If you do not respond within 30 seconds, the trial is recorded as a lapse and the test moves on, matching the standard PVT timeout
Reading your results on this test:
Mean RT is your average speed across all trials in the session
Lapses matter more than average speed — in research, lapse count rises sharply with sleep loss while average speed changes more slowly
False starts suggest you were guessing the timing rather than waiting for the signal — they do not speed up your average, because they are never counted as reaction times
Comparing your own sessions across different times of day is more informative than comparing yourself to anyone else
Performance commonly drifts slower across a longer session — that drift is the fatigue signal the task is built to expose
Screen type, input device, and browser all add a few milliseconds, so treat the absolute number as approximate