CS2 Mouse Polling Rate: Is 8000 Hz Worth the Cost
A reader sent me a clip in October of his frame rate counter during a smoke heavy retake on Mirage. It sat around 410 for most of the round, then dropped to 185 the moment two flashes popped and four players crossed mid.

A reader sent me a clip in October of his frame rate counter during a smoke heavy retake on Mirage. It sat around 410 for most of the round, then dropped to 185 the moment two flashes popped and four players crossed mid. He had bought an 8000 Hz mouse that same week and wanted to know why the upgrade had made his game feel heavier. The mouse was doing exactly what the box promised. His processor was paying for it.
Polling rate is the only mouse spec I know of that charges you rent. Every report the sensor sends is an interrupt the CPU has to service, and at 8000 reports per second you have multiplied that work eightfold over an ordinary 1000 Hz mouse. On a current generation chip with headroom to spare you will never see the bill. On the hardware most of us actually queue on, it shows up in the exact rounds where frame time matters most.
What the number buys you, in milliseconds
A mouse at 1000 Hz reports once every millisecond. Your click or your flick lands at some random point inside that window, so on average you wait half of it before the game hears about the movement. That is 0.5 ms of delay you cannot train away.
At 8000 Hz the window shrinks to 0.125 ms, and the average wait falls to about 0.06 ms. So the honest figure for the upgrade is roughly four tenths of one millisecond. For comparison, one frame on a 144 Hz panel takes 6.9 ms, and the slowest VSync setting I have measured added 38 ms by itself.
Four tenths of a millisecond is real. It is also about one twentieth of what you lose by playing a round at 90 frames instead of 200. That ratio is the whole argument, and it is why the first thing I ask anyone shopping for a high polling mouse is what their one percent low frame rate looks like on their worst map.
The cost nobody prints on the box
I run the same test on every mouse that comes through my desk now. Same map, same bot script, same 90 second loop, frame times logged, one pass per polling rate. These are the averages from a Ryzen 5 3600 paired with a 3060 Ti, which is roughly the machine that half the CS2 population plays on.
| Polling rate | Report interval | Average latency saved vs 1000 Hz | Average frame rate on my test rig | One percent low |
|---|---|---|---|---|
| 500 Hz | 2 ms | Costs you 0.5 ms | 398 | 176 |
| 1000 Hz | 1 ms | Baseline | 394 | 174 |
| 2000 Hz | 0.5 ms | 0.25 ms | 389 | 170 |
| 4000 Hz | 0.25 ms | 0.37 ms | 371 | 152 |
| 8000 Hz | 0.125 ms | 0.44 ms | 341 | 119 |
Look at the last column rather than the average. Going from 1000 Hz to 8000 Hz cost me 13 percent of average frame rate and 32 percent of the one percent low. In time terms, the one percent low fell from 5.7 ms per frame to 8.4 ms. I spent 0.44 ms to buy back 2.7 ms of occasional stutter, which is a trade I would not take at any price.
A newer chip changes the maths. On a Ryzen 7 7800X3D the same test lost 2 percent average and 4 percent on the low, which is close enough to measurement noise that I would not argue about it. If your processor is from 2022 or later and you play at 1080p with the GPU as your limiter, 8000 Hz is close to free. The problem is that the people most tempted by the spec are usually the ones least able to absorb it.
The counter take: 2000 Hz is where I stop
Ask in any CS2 Discord and the answer comes back fast. Latency is latency, max the slider, there is no downside. I have tested peripherals for six years and I think that advice is wrong for most of the people repeating it.
Two thirds of the available latency gain lives between 1000 Hz and 2000 Hz, and that step cost me 1 percent of average frame rate on the old rig. The last two doublings deliver 0.19 ms between them and take a bite out of frame time pacing that you can actually feel on a wide peek. I have settled on 2000 Hz on every mouse I own, including the ones that go to 8000, and when I ran a blind swap on four teammates at ranks between 18k and 23k Premier, none of them picked the 8000 Hz setting out of three options.
There is a quieter reason the spec sells well. A mouse that polls faster also reports smaller movements more often, which smooths out the stair stepping you get on slow tracking at low sensitivity. That part is genuine, and if you play at 400 DPI with a 50 cm turn you may prefer the texture of 2000 Hz over 1000 Hz for reasons that have nothing to do with latency. It just is not the reason the marketing gives you, and it tops out well before 8000.
The wireless side has its own penalty. A mouse rated for 90 hours at 1000 Hz typically gives 20 to 25 hours at 8000 Hz, because the radio wakes eight times as often. One of my daily drivers drops from 110 hours to 19. That turns a mouse you charge monthly into one you charge twice a week, and a flat mouse mid match is a worse outcome than 0.44 ms.
Testing it on your own machine in fifteen minutes
None of my numbers transfer to your setup, so measure yours. Load an offline server with bots on the map where your frame rate struggles most, not the one where it looks best. Set your mouse to 1000 Hz in its software, run a fixed route for 90 seconds with cl_showfps 2 open, and write down the average and the worst dip. Close the game, change only the polling rate, and repeat. Three passes per setting, because the first run after a launch is always slower.
If your one percent low falls by more than 10 percent at the higher setting, your CPU is the limiter and the polling rate is costing you more than it returns. Drop back a step. If nothing moves by more than 3 percent, you have the headroom and can run whatever you like.
Check the receiver placement before you blame the setting. A high polling dongle on a monitor hub or a front panel port produced irregular jitter on two of the three mice I tested that way, and the symptom reads exactly like a CPU problem. My walkthrough on testing CS2 input lag with a phone camera covers the measurement side if you want an end to end number rather than a frame rate graph.
What to change this week
Set your mouse to 2000 Hz, play four Premier matches, and leave it alone. That single change captures most of what the spec has to offer on any machine, and it is reversible in ten seconds if you hate it. Then put the money you were going to spend on a high polling mouse toward whatever is actually holding your frame rate floor down, which for most readers who write to me is 16 GB of memory running in single channel or a cooler that gave up in August.
One habit worth keeping: mouse software resets polling rate after some firmware updates, so confirm the value after every update rather than assuming it stuck. I have found mine quietly back at 1000 Hz twice this year. If you are also fiddling with sensitivity in the same session, change one variable at a time, and the method in finding your real cm per 360 will keep the two from contaminating each other. The rest of my peripheral notes sit in the gear section.
FAQ
Does 8000 Hz polling make you aim better in CS2?
Not in any way I have been able to measure. It removes about 0.44 ms of input delay compared with 1000 Hz, which is smaller than the variation between your own good and bad days. If the setting costs you frame rate stability, it will make your aim slightly worse rather than better.
What polling rate do professional players use?
Mostly 1000 Hz, with a growing minority at 2000 Hz and 4000 Hz. Pro configs are not a great guide here because tournament machines have far more CPU headroom than a home rig, so what works on stage may not survive on yours.
Will a higher polling rate drain my wireless mouse faster?
Yes, and the effect is large. Expect roughly a quarter of the rated battery life at 8000 Hz versus 1000 Hz. Running 2000 Hz usually costs between 30 and 40 percent of your runtime, which most people can live with.
Is 500 Hz ever the right choice?
Only on a very old processor that is already dropping frames at 1000 Hz. It adds half a millisecond of delay and I would treat it as a last resort rather than a tuning option.
More in Gear
Gear
Do You Need a Mouse Bungee for CS2 Aim Consistency
You hold the A site on Mirage, flick left onto a rushing player, and the shot lands a few pixels wide.
Gear
Mouse Grip Tape for CS2 and When It Actually Helps
Your mouse slips a few millimetres during a tense retake, your crosshair drifts off the head you were holding, and you lose the...
Gear
Is a Wireless Mouse for CS2 Worth It Over a Wired One
You are about to spend 80 to 160 dollars on a wireless mouse for CS2 and a voice in your head keeps asking if you will lose...