Input lag

Input lag is the delay between your action, or a change on screen, and the moment it registers: the time your mouse, computer and screen add on top of your own reaction. A faster screen helps mainly because it lowers that delay. In a study of skilled esports players, cutting the delay improved their aiming clearly, while a higher refresh rate on its own, with the delay held equal, made little difference^[1]. A reaction test in a browser always includes some input lag: robot tests of ordinary computers and phones found delays of about 57.6 ms to 132.9 ms^[2].

Test your screen's refresh rate

The figures in one table

Delays added by equipment, in the studies on this page
What was measured	Figure
A tuned gaming computer, click to light on a 60 Hz screen, lowest setting	22 ms
The same computer with delay added on purpose	55 ms and 88 ms
A calibrated laboratory screen and mouse: reaction time before and after removing their delay	230.8 ms and 213 ms
A browser on a phone or laptop, robot press to recorded time	57.6 ms to 132.9 ms
Common browsers on desktop and laptop computers, the same kind of measure	76.5 ms to 82.3 ms

The browser figures are larger than the gaming ones because they include everything from the press of a key or screen to the moment a web page records it, on ordinary devices and settings. They were measured with robots, so no human reaction is in them.

Where the delay comes from

The input device. A mouse or keyboard checks its buttons and sends a report to the computer at a fixed rate; a touch screen needs time to notice a finger.
The computer. The operating system passes the input to the browser or game, which works out what to draw next.
The screen. A new picture is shown only at the next refresh, and the panel then takes a moment to change its pixels.

Each step is small, but they add up, and the same chain works in the other direction for a reaction test: the signal waits for the screen, and your press waits for the device and the browser.

What refresh rate does

A 60 Hz screen draws a new picture sixty times a second, so a change can wait up to a sixtieth of a second before it appears. A 144 Hz screen redraws 144 times a second, so the worst wait falls to a hundred-and-forty-fourth of a second. That is a real saving of a few thousandths of a second, and motion also looks smoother.

Researchers at a graphics company asked which matters more: the higher refresh rate itself, or the lower delay that comes with it. They tested eight skilled esports players aged 27 to 36 at 60, 120, 240 and 360 Hz, and added delay at the higher rates so that the delay was the same at every rate. Aiming at targets got clearly faster when the delay was lower. Refresh rate alone had no significant effect on hitting a target with one click, and only a small one on following a moving target^[1].

Their title sums it up: a delay of 30 ms matters more than a refresh rate above 60 Hz. A fast screen is worth having mainly because it usually comes with less delay.

Mice and keyboards

A mouse reports to the computer a fixed number of times a second, its report or polling rate. One maker states 125 (as the maker states)^[4] reports a second as the slowest setting of a gaming mouse and 1,000 (as the maker states)^[4] as its fastest. A 1000 Hz mouse reports eight times as often as a 125 Hz one, which shortens the wait for the next report by a few thousandths of a second at most.

Keyboards differ more than people expect. The robot study of browsers found a laptop with a built-in keyboard no slower than a desktop with a plug-in one: on Windows, the laptop averaged 73.7 ms and the desktop 76.2 ms^[5]. The whole set-up mattered more than the type of keyboard.

What a browser adds

Two studies pressed devices with robots and compared the real time of each press with the time a web page recorded. On two laptops and two phones that were a few years old at the time, every browser recorded reactions as slower than they were, by an average of 57.6 ms on an iPhone in Safari up to 132.9 ms on a Mac in Safari^[2]. On desktop and laptop computers, the second study found average delays of 76.5 ms to 82.3 ms across browsers^[5].

Precision is a separate question from delay. A delay that is the same every time shifts every result equally and does not change a comparison between two of your own runs; a delay that varies blurs small differences. A large comparison of experiment software found that most packages measured response times in browsers with a precision under 10 ms, against under 1 ms for the best laboratory software^[6].

The robot figures include the whole chain on everyday hardware, not only the browser. They show why a reaction time measured on a website reads higher than one from a laboratory, and why results from two different devices should not be compared as if they were the same instrument.

What this means for your own result

A faster screen or mouse can shave a few thousandths of a second off your measured time; your own reaction is usually ten times that or more.
The biggest gain is lower overall delay: a wired mouse, a screen set to its highest refresh rate, and other programs closed.
Results from different devices differ by more than a better screen can change. Compare your results on one device.

Our reaction tests allow for the delays these robot studies found, less the delay the laboratory study's own equipment had: 44.1 ms to 115.1 ms with a keyboard or mouse, and 39.8 ms to 52 ms with a touch screen. No study pressed a mouse, so a mouse gets the keyboard's figures. The comparison is shown as a range for that reason, and on a computer the range is wide, so treat it as a rough guide. The refresh rate test and the polling rate test show what your own screen and mouse are doing.

How laboratories, sprint starts and browsers each time a reaction is explained in how reaction time is measured.

Questions

Does a 144 Hz screen make you react faster?

Not your own reaction. It can lower the delay of the screen a little, which lowers your measured time a little. In a study of esports players, lower delay helped aiming clearly, while refresh rate on its own, with delay held equal, made little difference.

How much input lag does a browser add?

On the everyday laptops and phones tested with a robot, the recorded times were too slow by an average of 57.6 ms to 132.9 ms, depending on the device and browser.

Is a 1000 Hz mouse worth it?

It shortens the wait for the next report by a few thousandths of a second. That is real but small compared with the rest of the chain and with a human reaction.

Why is my reaction time different on another computer?

Because each device adds its own delay. Robot tests found tens of thousandths of a second of difference between browsers and devices, so compare your results on one device.

Sources

Spjut, J., Boudaoud, B., Binaee, K., Kim, J., Majercik, A., McGuire, M., Luebke, D., & Kim, J. (2019). Latency of 30 ms benefits first person targeting tasks more than refresh rate above 60 Hz. SIGGRAPH Asia 2019 Technical Briefs.doi:10.1145/3355088.3365170
Eight skilled esports players aged 27 to 36
Title: "Latency of 30 ms Benefits First Person Targeting Tasks More Than Refresh Rate Above 60 Hz". Abstract: "we isolate latency and refresh rate by artificially increasing latency when operating at high refresh rates. Eight skilled esports athletes then perform gaming-inspired first person targeting tasks under varying conditions of refresh rate and latency" "We show that reduced latency has a clear benefit in task completion time while increased refresh rate has relatively minor effects on performance when the inherent latency reduction present at high refresh rates is removed." "Subjects. Eight subjects (aged 27-36)" "The selected refresh rates were: 60 Hz (today's baseline), 120 Hz (today's gamer), 240 Hz (gaming enthusiast) and 360 Hz (experimental)." Table 1, click to photon latency at 60 Hz: "22 (6.0) 55 (5.9) 88 (5.8)". Figure 1: "For the 1-hit tasks, latency has a significant effect on task completion times while the effect of refresh rate is not statistically significant." (read on 2026-10-06: https://research.nvidia.com/sites/default/files/pubs/2019-11_Latency-of-30/FPS_Tasks_sa2019_AuthorVersion.pdf)
Pronk, T., Wiers, R. W., Molenkamp, B., & Murre, J. (2020). Mental chronometry in the pocket? Timing accuracy of web applications on touchscreen and keyboard devices. Behavior Research Methods, 52(3), 1371–1382.doi:10.3758/s13428-019-01321-2
Four devices from 2015 and 2016: a MacBook Pro, an ASUS laptop, a Samsung Galaxy S7 and an iPhone 6S, pressed by a robot
Abstract: "In controlled circumstances, as can be realized in a lab setting, very accurate stimulus timing and moderately accurate RT measurements could be achieved on both touchscreen and keyboard devices, though RTs were consistently overestimated." Table 4 "Descriptives of RT overestimations (in milliseconds) per device and browser" (OS, Web Browser, Minimum, Maximum, Mean, SD): "Android Chrome 46.0 103.5 69.8 7.4", "iOS Safari 48.3 96.3 57.6 6.5", "MacOS Safari 93.0 163.7 132.9 8.1", "Windows Chrome 64.7 70.6 68.5 1.7", "Windows Firefox 49.8 84.9 61.9 5.7". (read on 2026-10-06: https://www.ebi.ac.uk/europepmc/webservices/rest/PMC7280355/fullTextXML)
Woods, D. L., Wyma, J. M., Yund, E. W., Herron, T. J., & Reed, B. (2015). Factors influencing the latency of simple reaction time. Frontiers in Human Neuroscience, 9, 131.doi:10.3389/fnhum.2015.00131
n = 1,469; Community volunteers in Rotorua, New Zealand, aged 18 to 65 (mean age 45.8, 40% men)
Abstract: "Experiment 1 examined a community sample of 1469 subjects ranging in age from 18 to 65. Mean SRT latencies were short (231, 213 ms when corrected for hardware delays) and increased significantly with age (0.55 ms/year)". Table 2, column Experiment 1: "N 1469", "SRT 230.8", "SRT SD 26.8". (read on 2026-10-01: https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2015.00131/full)
Logitech (2026). Customize G PRO gaming mouse pointer settings with Logitech Gaming Software. Logitech Support.https://support.logi.com/hc/en-us/articles/360023183294-Customize-G-PRO-gaming-mouse-pointer-settings-with-Logitech-Gaming-Software
A gaming mouse and its settings, as the maker describes them
"Change the Report Rate, if you prefer something other than the default of 1000 reports/second (1ms response time). The Report Rate controls how often your mouse updates its position to the computer. NOTE: The possible range is from 125/sec (8ms response time) to 1000/sec (1ms response time)." (read on 2026-10-06: https://support.logi.com/hc/en-us/articles/360023183294-Customize-G-PRO-gaming-mouse-pointer-settings-with-Logitech-Gaming-Software)
Anwyl-Irvine, A., Dalmaijer, E. S., Hodges, N., & Evershed, J. K. (2021). Realistic precision and accuracy of online experiment platforms, web browsers, and devices. Behavior Research Methods, 53(4), 1407–1425.doi:10.3758/s13428-020-01501-5
Desktop and laptop computers with Windows 10 and macOS, pressed by a robot; and the equipment of 202,600 online participants
Abstract: "We then employed a robot actuator in realistic set-ups to measure response recording across the aforementioned platforms, and between different keyboard types (desktop and integrated laptop)." "We found that modern web platforms provide reasonable accuracy and precision for display duration and manual response time". Table 2 "RT delay is calculated as the difference between known and recorded RT." Browser means: "Chrome 78.81", "Edge 80.10", "Firefox 82.30", "Safari 76.50"; device means: "macOS-Desktop 85.35", "Windows-Desktop 76.24", "Windows-Laptop 73.65". Results: "We found that 77% of these devices were desktop or laptop computers, whereas only 20% were mobile devices" "Based on a sample of 202,600 participants." (read on 2026-10-06: https://www.ebi.ac.uk/europepmc/webservices/rest/PMC8367876/fullTextXML)
Bridges, D., Pitiot, A., MacAskill, M. R., & Peirce, J. W. (2020). The timing mega-study: comparing a range of experiment generators, both lab-based and online. PeerJ, 8, e9414.doi:10.7717/peerj.9414
Abstract: "Among the lab-based experiments, Psychtoolbox, PsychoPy, Presentation and E-Prime provided the best timing, all with mean precision under 1 millisecond across the visual, audio and response measures." "Online studies did not deliver the same level of precision as lab-based systems, with slightly more variability in all measurements." "For response times (measured using a high-performance button box), most of the packages achieved precision at least under 10 ms in all browsers, with PsychoPy achieving a precision under 3.5 ms in all." "The results, from over 110,000 trials, highlight the wide range of timing qualities that can occur even in these dedicated software packages for the task." (read on 2026-10-06: https://pubmed.ncbi.nlm.nih.gov/33005482/)