Wait for the screen to turn green, then click or tap as quickly as you can. Your result is the average of five recorded trials.
Click the button below or tap anywhere to start.
Your five measured responses (lower time is faster).
Results are browser-based estimates and can vary with display, input device and system latency. They are not laboratory measurements.
A reaction time test measures the delay between a signal appearing and your response to it. In this free online visual reaction time test, the signal is a screen changing from red to green, and the response is a click, tap or keyboard activation. The result is measured in milliseconds (ms), where 1,000 milliseconds equal one second. A lower number means a faster recorded response for that trial.
This is a simple visual reaction task: you already know what action to take and only need to recognize when to take it. It is different from a choice reaction test, where you must identify a signal and select between several possible responses. ReflexBenchmark makes the test quick to repeat and shows the five measurements behind your final score, rather than hiding them behind a single number.
If you press while the screen is still red, the test counts a false start. That attempt is not added to your five valid reaction times; you retry the same trial. Switching browser tabs or minimizing the window while waiting also interrupts the current trial instead of counting a misleading measurement.
Your main score is the arithmetic mean of five valid visual responses. The browser measures each response with a high-resolution timer, rounds each trial to whole milliseconds and calculates the average of those five values. The displayed final average is rounded to the nearest millisecond. A low result is faster, but it still includes the delay contributed by your screen, input device, browser and operating system.
Suppose your five trials are 230, 245, 220, 255 and 250 ms. Their total is 1,200 ms. Dividing by five gives an average reaction time of 240 ms. The best response is 220 ms, the slowest is 255 ms, and the middle value after sorting is a median of 245 ms.
These are example calculations, not a claim about the average ReflexBenchmark user.
The result panel also displays standard deviation, a measure of how spread out your five times are. The graph shows each measured trial separately. False starts are shown as a count but are not included in the score average. When enough comparable recorded attempts exist, the site may also report an estimated percentile; otherwise, the rank remains pending.
A single online result is best understood as a snapshot of performance in that particular session and setup. Two people can produce different browser scores even if their underlying response abilities are similar, because devices register and display events at different speeds. A number such as 240 ms means that the browser recorded an average delay of 240 milliseconds over five accepted trials. It does not establish a medical diagnosis, a permanent personal ability or a verified population ranking.
Look at the average together with the trial spread. Five tightly grouped responses suggest more consistency in that session than a mix of very fast and very slow clicks. If your best and worst differ substantially, attention, anticipation, interruptions or device timing may have influenced the result. Compare repeated sessions on the same setup before interpreting a small change as meaningful.
If a percentile is displayed, it compares the recorded score with other qualifying website attempts using the site's available data. Those attempts are not a representative sample of all people, and a percentile should not be described as an age-based clinical norm. A pending rank simply means the system has not established a comparison it can report.
Display and input latency: Monitors, touchscreens, mice, keyboards and browser event processing add delays between what is shown, what you do and what software records. Their effects differ by model and configuration. A touchscreen score and a desktop mouse score are therefore not interchangeable measurements.
Attention and anticipation: Distracted responses may be slower. Guessing when the screen will change can produce unusually short times or false starts, neither of which provides a reliable picture of normal visual response. The randomized waiting interval is intended to make a simple fixed-countdown strategy less useful.
Practice and familiarity: Knowing where to click and becoming comfortable with the sequence can change how you perform across sessions. Sleep, fatigue, temporary stress and general alertness may also affect performance. Avoid assuming that one specific habit or supplement guarantees a fixed number of milliseconds of improvement.
Research on remote cognitive testing has found that personal digital devices can influence measured scores. For additional context, see the open-access study Cognitive test scores vary with choice of personal digital device.
If your goal is practice rather than a personal benchmark, you can use several sessions to learn the interaction. If your goal is comparison, aim for consistent conditions and avoid cherry-picking only your fastest attempt.
ReflexBenchmark uses browser timing, not laboratory equipment. The timer begins around the frame in which the green cue is prepared, so display refresh timing and hardware response are not independently calibrated. The recorded interval includes parts of visual detection, decision to respond, movement and device event handling. It is useful for informal practice and self-comparison, not for diagnosing neurological conditions or measuring an exact neural conduction time.
The server accepts scores under its validation rules, but a saved score is not proof that a person completed the test unaided. The trial chart reflects values reported by the browser during the attempt. If the score API cannot confirm saving, your local result can still appear, but the website should not claim it was saved successfully.
To compare a different type of response, try the Audio Reaction Time Test for an auditory signal or the Choice Reaction Time Test for selecting between possible responses. The Aim Trainer Test adds cursor targeting, while the Processing Speed Test focuses on fast visual matching. These tasks measure related but distinct performance, so their scores should not be compared as if they used the same scale.
The test records five valid responses to the green screen. The final score is the rounded average of those five times.
That is a false start. It increases your false-start count, but the attempt is not added to the five measured trials. You can retry that trial.
The browser measures each valid response and rounds it to whole milliseconds. The displayed score is the mean of all five valid trials, rounded to the nearest millisecond.
Yes. For this specific task, fewer milliseconds means the browser recorded a faster response. Also compare your best, worst and trial-to-trial variation.
Touch input, mouse or keyboard input, display refresh, and device latency can differ. Use the same hardware and input method when tracking your progress.
A comparison is shown only when the site has enough eligible recorded attempts. A pending rank does not mean your result is invalid or that you performed poorly.
No. It is an informal online visual response exercise affected by browser and device timing. It is not a clinical assessment or a substitute for professional evaluation.