Input Lag: How to Test Keyboard and Mouse Latency

Input lag is the delay between pressing a key or clicking and seeing the result on screen. Searching for a keyboard latency test or a mouse latency test usually turns up websites that promise to measure it in your browser. They cannot measure your device’s real latency, and this guide explains why, how reviewers and NVIDIA actually measure it, and what you can do to reduce input lag.

What input lag is made of

NVIDIA, which builds latency measuring tools into its gaming displays, avoids the vague term “input lag” and splits the full delay, called system latency, into three parts:

Part	What it covers
Peripheral latency	Your mouse or keyboard processing the press and sending it to the PC
PC latency	The operating system, the game and the rendering of a new frame
Display latency	The monitor showing the new image

NVIDIA defines system latency as “the time it takes your mouse clicks to end up as pixels on screen”, and says PC latency “is often the largest contributor”. The device itself is only one part: NVIDIA says mice and keyboards “can range anywhere from 1ms of latency to ~20ms of latency”.

Can a website test keyboard or mouse latency?

Not the latency of the device itself. A web page only learns about a key press or click when the browser creates an event for it, after the mouse or keyboard, the USB connection and the operating system have already done their part. The page cannot see when your finger pressed the switch, and it cannot see when the screen changed.

Browsers also make their timers deliberately less precise. MDN, the main web developer reference, says the event timestamp’s “precision may be reduced to mitigate timing attacks and fingerprinting”. In Chrome it is rounded to 0.1 ms, in Safari and Firefox to 1 ms, and in Firefox with fingerprinting protection turned on to 16.667 ms. Device latency differences are often only a few milliseconds, so this rounding matters.

What a browser test can show:

The time between two events it receives, such as the gap between two clicks on the double click test.
Whether every press arrives, and arrives once, on the mouse test.
How often the mouse reports movement, on the polling rate test.
Your own reaction time on the reaction time test, which includes your whole system’s delay plus your response. An RTINGS tester notes that reacting to something on screen will “almost certainly” take over 100 ms already, far more than the device adds.

How mouse and keyboard latency is really measured

Reviewers: a robot finger and a USB analyzer

RTINGS, which publishes test results for hundreds of mice and keyboards, measures latency with hardware:

Keyboards: a solenoid presses the key, and a Beagle 480 USB analyzer intercepts the data on its way to the computer. RTINGS takes 200 samples and measures from the moment the solenoid touches the key, so the result covers the keyboard only, not the operating system or the game.
Mice: the same kind of setup clicks the button 205 times and averages the results. The measurement includes the button travel before the switch actuates, so RTINGS says its numbers are typically slightly higher than manufacturers’ figures.

RTINGS rates mouse click latency under 10 ms as good for wired mice and wireless mice with a USB receiver, and under 20 ms as good for Bluetooth.

NVIDIA Reflex Latency Analyzer

Some 360 Hz G-SYNC displays include NVIDIA’s Reflex Latency Analyzer. You plug the mouse into a USB port on the monitor, which passes the clicks through to the PC and notes when each one happens. The monitor then measures when the pixels change on screen, for example a muzzle flash. It splits the result into mouse latency, PC plus display latency and full system latency. Any mouse works for the PC and display part except Bluetooth mice; measuring the mouse itself needs a compatible model.

Before tools like this, NVIDIA says the usual method was a 1000 FPS high speed camera, a setup that “starts at about $7,000 USD”.

What affects mouse and keyboard latency

Polling rate. This is how often the computer asks the device for new data. At 1,000 Hz that is every 1 ms, at 8,000 Hz every 0.125 ms. NVIDIA says a low rate of 125 Hz “adds up to 3ms of system latency on average compared to a 1000Hz polling rate”. Read our guide to polling rate for the trade-offs.
Debounce time. Mechanical switches wait a few milliseconds after each press so contact bounce is not read as a second click. Shorter debounce means a faster response, but too short can cause double clicking.
Connection. RTINGS says wired mice and wireless mice with their own USB receiver generally have the lowest latency, while Bluetooth mice have the highest. Logitech says its LIGHTSPEED receivers offer “faster response times and more stable connections compared to Bluetooth”.
Receiver distance. RTINGS notes that USB receivers tend to add latency the further away the mouse is, so keep the receiver close.
Not DPI. NVIDIA is clear: “Higher DPI does not mean lower latency.”

Keyboard delayed input: check Filter Keys

If key presses arrive late or some are ignored, check a Windows accessibility feature before blaming the keyboard. Filter Keys lets Windows “ignore brief or repeated keystrokes”. Its settings include a delay before a key is accepted at all and a gap during which a second press of the same key is ignored, so typing can feel delayed or missing.

It is easy to turn on by accident: holding the right Shift key for eight seconds switches Filter Keys on or off.

To check it:

Press Windows logo key + U, or select Start > Settings > Accessibility.
Select Keyboard.
Turn the Filter keys switch off.

If held keys repeat too slowly, change the character repeat delay. In recent Windows 11 updates Microsoft moved this setting from Control Panel to Settings > Accessibility > Keyboard; older builds keep it in Control Panel under Keyboard.

How to reduce input lag

Use a wired connection or the mouse’s own USB receiver instead of Bluetooth, and keep the receiver close.
Set the polling rate to at least 1,000 Hz in the mouse software. Logitech notes that rates above 1,000 Hz can cause frame rate drops on some systems, so test before keeping a higher one.
Lower the debounce time only as far as you can without double clicks.
Turn off Filter Keys if you do not use it.
In games, use any low latency mode the game offers and keep the frame rate high, because PC latency is often the biggest part.

Frequently asked questions

Can I test my keyboard latency online?

Not accurately. A website only sees key events after the keyboard, USB and operating system have delivered them, and browsers round their timers. Real keyboard latency is measured with a solenoid and a USB analyzer, as reviewers like RTINGS do.

What is a good mouse latency?

RTINGS rates click latency under 10 ms as good for wired mice and mice with a USB receiver, and under 20 ms for Bluetooth mice.

Why is my keyboard input delayed?

A common cause in Windows is Filter Keys, which can make the keyboard ignore brief or repeated presses. Turn it off in Settings > Accessibility > Keyboard. Wireless connection problems and low batteries can also delay input.

Does polling rate reduce input lag?

Yes, up to a point. NVIDIA says 125 Hz adds up to 3 ms on average compared with 1,000 Hz. Above 1,000 Hz the gain is under 1 ms per report, and very high rates can cost frame rate on some PCs.

Is Bluetooth worse for latency?

Generally yes. RTINGS says Bluetooth mice have the highest latency, and Logitech says its own receiver connection is faster than Bluetooth. For games, use a wired connection or the mouse’s USB receiver.

Does a higher DPI lower input lag?

No. NVIDIA states that higher DPI does not mean lower latency, so choose the DPI that suits your aim.