Reaction Time Test

How to use

1. Click or tap the large test panel, or use Start, to begin.
2. Wait for the panel to turn green. An early response counts as a false start and does not count as a valid sample.
3. Click or tap the panel, press Space, or press Enter while the panel is focused as soon as it turns green.
4. Complete five valid attempts, then review average, median, best, worst, false starts, response input, and local history.

What the result means

• Each valid sample is the elapsed browser-observed interval from the timestamp used to trigger the green cue to the pointerdown or keyboard response received by the page.
• Lower times are faster. Average is the mean of five valid samples; Median is the middle valid sample and is less affected by one unusually slow attempt.
• Best and Worst show the spread of the session. Treat the Best attempt as one sample, not as your overall reaction time.
• Response input records the browser-reported method for each valid attempt. A Mixed session remains valid, but it is less useful for direct comparison with a single-method session.
• The result still includes the browser/rendering/display path before you perceive the cue, your response, the input device, and browser event delivery.
• For meaningful comparisons, repeat five-attempt sessions using the same display, browser, response method, similar system load, and the same wired or wireless input setup where relevant.
• Local best average and session history are stored only in this browser.

Limitations

• The timer starts when the page triggers the green cue inside a browser animation frame; DeviceMetrik does not use a physical sensor to detect the exact instant a display pixel changes.
• Results include human response plus display, browser, operating-system/input-path, and input-device delay. This is not pure mouse latency, keyboard latency, display latency, or a laboratory measurement.
• Refresh rate, mouse vs touch vs keyboard input, browser scheduling, device load, power-saving behavior, and related system conditions can shift the measured result.
• Backgrounding or losing focus aborts a running attempt rather than recording it as a valid reaction.

Troubleshooting

Clicks do nothing after Start

Click or tap the large reaction panel, or use Start below it. Wait for the green signal; an early response is a false start.

Test stops unexpectedly

Leaving the tab or unfocusing the window aborts a running trial for accuracy.

My results vary a lot between sessions

Use the same display, browser, response method, and wired or wireless input setup; keep the page focused; avoid heavy background workload; compare repeated five-attempt sessions; inspect the median as well as the average.

Related tools

• F1 Reaction Time Test

  Run an unofficial five-light F1 reaction time test with randomized lights out. Compare five valid starts, input method, false starts, median, and best time.

• Keyboard Tester

  Test keyboard keys online across common Windows, Mac, and compact layouts. See pressed and detected keys, event.key/event.code, repeat state, simultaneous browser input, and untested keys.

• Mouse Test

  Use this online mouse tester to check buttons, side buttons, held states, wheel events, and pointer movement. See browser counts, hold times, movement path, and session status.

FAQ