By Haru · Updated · 2026-08-30
What DPI Should I Use — The Number Alone Decides Nothing
Two completely unrelated things are called DPI, and search results mix them freely. One is about mice: how many steps the sensor reports per inch of movement. The other is about print and scanning: how many dots of ink or samples per inch. Nothing said about one applies to the other. This page is about the mouse.
The usual answer to the question — 800, or 1600, or whatever number a forum settled on — is not wrong so much as incomplete. DPI is one factor in a chain, and changing it while leaving the rest alone moves the thing you actually feel by an amount nobody can predict from the DPI number alone.
What you can compare between two people, two mice and two games is how far your hand has to travel to turn all the way around. That distance is a real measurement, it is the same on any setup, and once you know yours the DPI question mostly answers itself.
So: what the number does and does not control, the arithmetic that connects it to your hand, and how to pick a value without copying someone else's.
Two different DPIs, and only one is about your mouse
Search for a DPI recommendation and half the results are about images. That is not an accident: the same three letters got attached to two unrelated measurements decades apart.
Print and scanning DPI counts dots or samples per inch of paper, and it is about how much detail is captured or laid down. It is why a scan at 300 is fine for a document and 600 is asked for on photographs, and it has no relationship to any input device.
Mouse DPI counts how many movement steps the sensor reports for every inch the mouse travels. Nothing is being resolved or printed; it is a rate of reporting distance. Some manufacturers prefer the term CPI, counts per inch, for exactly this reason, and it is the more honest name.
If an article about DPI mentions paper, resolution or image quality, it is answering the other question. Close it and keep looking. Everything below is about the mouse and nothing on this page will help with a scanner.
Why the number on its own decides nothing
A mouse DPI setting is one multiplier in a chain, and the chain is what your hand feels.
The sensor reports counts. The operating system may scale them, and a game applies its own sensitivity multiplier on top. The result is how far your view turns for a given hand movement. Two people on 800 DPI can be twice as far apart as two people on 400 and 1600, because the sensitivity settings sitting after the DPI are doing more of the work than the DPI is.
Copying a DPI number from a forum so rarely feels right because you inherited one link in the chain and kept your own values for the rest, so the thing that actually changed is not the thing you were trying to match.
The same chain is why raising DPI does not make you faster in any useful sense. It makes the pointer or the view move further for the same hand movement, which you can achieve just as well from the sensitivity slider. What it changes on its own is only how finely that movement is divided.
The number that can actually be compared
There is one figure that survives being carried between setups: how far the mouse has to move for a full turn.
It is usually written as centimeters per 360, and it is exactly what it sounds like — the distance your hand travels to spin all the way around. Because it is a physical distance it means the same thing on any mouse, at any DPI, in any game. Two people with the same cm/360 have the same feel even if none of their settings match.
Getting yours takes about a minute and no maths. Put the mouse against something on the desk, turn exactly one full revolution in a game, mark where the mouse ended up, and measure the gap. Write it down. That number is now your setting, and any DPI or sensitivity change that keeps it is a change you will barely notice.
The conversion is simple when you do want it: multiplying DPI by in-game sensitivity gives a combined figure that halves the distance when it doubles. Doubling DPI and halving sensitivity leaves the distance where it was — which is the whole trick to changing DPI without relearning your aim.
What actually changes when you go higher or lower
Set aside speed, which the sensitivity slider controls anyway, and a few real differences remain.
Higher DPI divides movement more finely, so the smallest movement the mouse can report is smaller. On paper that means more precision. In practice a hand tremor gets reported just as faithfully as an intended movement, and at very high settings some people find the cursor feels unsettled for exactly that reason.
Lower DPI reports in coarser steps. Very low settings on very high resolution screens can leave the pointer unable to land on a specific pixel, which matters for detailed work and almost never for games.
The worry you can retire is pixel skipping. That was a real limitation of older optical sensors pushed past their tracking speed, and modern sensors track far faster than a hand moves. If a page warns you about it as a reason to avoid a setting, the page is old.
The genuinely useful rule is that the middle is uncontroversial. Settings between roughly 800 and 1600 avoid both ends, and the choice inside that range matters far less than what you multiply it by.
Choosing by what you are doing
Different work pulls in different directions, and the honest advice is a range rather than a value.
For everyday desktop use the constraint is your screen. On a high resolution display, a low DPI plus a low pointer speed means dragging across the desk to cross the screen. Raise it until one comfortable arm movement crosses the display and stop; there is nothing to optimize past that.
For detailed drawing or editing, precision is the point and speed is not. Many people work lower than their desktop setting deliberately, because a slow pointer is easy to place. If you are on a tablet the pen has its own settings and the mouse figure does not apply.
For games, the split is between styles rather than between titles. Aiming that relies on large, fast turns wants a shorter distance per revolution; aiming that relies on holding a small target wants a longer one. Both are reached by adjusting sensitivity around whatever DPI you picked, not by chasing a DPI number.
And for anything competitive, consistency beats optimization. A setting you have used for months will outperform a better one you adopted yesterday.
Changing it without losing your aim
The reason DPI changes feel awful is almost always that the distance changed with them, and that is avoidable.
Measure first. Get your current cm/360 before touching anything, because without it you have nothing to restore and you are guessing at whether the new setting is better or merely different.
Change DPI and sensitivity together in opposite directions. Double one, halve the other, and the distance stays put. What changes is only the fineness of the steps, which is what you were trying to adjust in the first place.
Check what the mouse is actually set to rather than what you believe it is set to. Onboard profiles get switched by a button that is easy to press by accident, and the step you are on is not always the one you configured. Measuring it takes a minute and settles the question.
Then leave it alone for a week. Every new setting feels wrong for a day or two, so a change judged on the first evening is judged on unfamiliarity rather than on merit.
Frequently asked questions
What number should I actually pick?
Anywhere between roughly 800 and 1600 is uncontroversial, and inside that range the choice matters far less than the sensitivity you multiply it by. Pick one, then set your sensitivity so a full turn takes a hand movement you can make comfortably. That distance is the setting; the DPI is just one of the two numbers that produce it.
Is a higher DPI mouse better?
Not in itself. A high maximum is a specification, not a benefit, and almost nobody runs near the top of the range on a mouse that advertises tens of thousands. What matters is that the sensor tracks reliably at the setting you actually use, which is true of essentially any current mouse in the middle of its range.
How do I find out what my mouse is set to now?
Measure it, because there is no way to ask the mouse. Our DPI test does it by counting the movement your mouse reports over a distance you measure with a ruler, which is the same method any honest page uses. Onboard profiles can also be switched by accident, so the value you measure is worth more than the value you remember configuring.
Does DPI affect polling rate, or the other way round?
Neither. Polling rate is how often the mouse reports, in reports per second, and DPI is how much movement each report describes. A mouse can be high in one and low in the other, and changing either leaves the other alone. If your movement feels stuttery rather than too fast or too slow, that is the polling rate question and not this one.
Do I need to change Windows pointer speed too?
For games, usually not, since most read the mouse directly and ignore the system setting. For desktop work the two multiply, so if you raise DPI and leave pointer speed high the result overshoots. The common advice to leave pointer speed at the middle notch and turn off pointer precision exists so that the DPI number means something predictable.
Is it worth copying a professional player's settings?
Only as a starting point, and only if you convert it properly. Their DPI without their sensitivity tells you nothing, and even the pair only transfers if the game is the same. What does transfer is the distance per turn, so convert their setup to that figure, try it, and then adjust — because their arm, desk and grip are not yours.