Stopwatch
The stopwatch above starts on a click and counts up until you stop it. It measures against the system clock, so a background tab keeps its time. Reset takes it back to zero.
0:00.00
Timed against the system clock rather than a frame counter, so it stays accurate if the tab is backgrounded.
Every page in this section
Counting up, not down
A stopwatch and a timer answer different questions. Use the stopwatch when the length is the thing you want to find out: how long the run took, how long the meeting actually ran, how long the glue needed. Use a countdown when you already know the length and want to be told when it is up. If you find yourself watching the stopwatch and waiting for a number, you wanted a timer.
Laps
A lap marks a point without stopping the clock, so you get the split and the running total at once. That is what you want for repeated efforts of the same kind – lengths of a pool, circuits of a track, one item after another on a production line. The stopwatch with laps keeps the list on screen so you can compare the splits afterwards instead of trying to remember them.
How precise a hand-timed number really is
The clock is accurate. The hand on it is not. You start it after the event begins and stop it after the event ends, and both of those gaps are your own reaction, which for most people sits somewhere around 200 to 300 milliseconds. The two errors partly cancel, but not reliably. What that means in practice:
- Over a long effort, a couple of tenths is noise. A 40-minute walk timed by hand is fine.
- Over a short one, it is most of the measurement. Hand-timing a 5-second sprint tells you very little.
- Comparing two runs you timed yourself is fairer than comparing your number to someone else’s, because the same delay is in both.
If you are curious how big your own share of that is, the reaction time test puts a number on it – though that number includes your screen, mouse and browser as much as you, and means nothing clinical.
Background tabs and sleeping phones
Elapsed time is read from the system clock rather than counted tick by tick, so switching tabs or locking the screen does not lose time: come back and the total is right. What can fail is anything that has to make a noise while you are not looking. For a long unattended run, use a device that rings on its own.
Elsewhere on the site
If what you want is a count rather than a duration, the click counter tallies presses with no clock attached, and the click speed test puts a fixed window around it.