Run a single downward movement and avoid rocking the wheel back. Repeat upward after a pause. If unwanted reversals recur, compare a second application and another mouse on the same computer. This separates an application-specific jump from a symptom that follows the device.
Part of the Mouse test. Run the full mouse test first for a complete device check.
The wheel jumps backwards Pages jump but no reversals appear Scrolling works but middle click does not
Mouse scroll test
Run a mouse scroll test to check direction, wheel events and unexpected reversals. Scroll steadily in one direction, pause, then test the other direction before interpreting the counts.
Also known as: mouse scroll test, scroll wheel test, scroll test, scrolling mouse test, mouse wheel test, scroll tester, scroll wheel tester. Free, online, and no download required.
No permission neededNothing uploaded
Move the pointer into the test area and roll the wheel in one direction.
01
What the mouse scroll test checks
| Check | How it is measured | Fails when |
|---|---|---|
| Wheel direction | Positive and negative vertical wheel events received by the page. | Events arrive in the opposite direction during a deliberate one-way movement. |
| Reversal count | Events opposing the majority direction within a continuous segment of at least four events. | Unexpected reversals recur in separate one-way runs. |
| Step size | The median absolute wheel delta, interpreted with its pixel, line or page mode. | Different delta modes make a direct size comparison invalid. |
How to run the mouse scroll test
- Keep the pointer inside the test stage and start a fresh run.
- Roll steadily downward without deliberately changing direction.
- Stop fully, then roll steadily upward as a separate movement.
- Finish the run and compare reversals with movements you actually made.
- Repeat in another application before deciding that the wheel needs repair.
If something failed
A page can move because content loads, focus changes or scrolling software intervenes. Compare a static document with this test. Zero reversals only describes the recorded wheel events; it does not rule out missed inputs or changes outside the sample.
Wheel rotation and pressing the wheel are separate inputs. Open the mouse button test, press and release the wheel, and check its button entry. Changing scroll speed will not explain a missing middle-button event.
02
How to read the result
How to run a scroll wheel test
Use a ratcheted wheel mode if your mouse offers one, and begin with ordinary steady scrolling rather than a free spin. The purpose is to compare your intended direction with the events received by the page. Do not alternate up and down during the same movement, because a deliberate reversal looks like an unexpected one to the counter.
A pause longer than 150 ms separates continuous segments in this implementation. A segment needs at least four non-zero vertical events before reversal detection applies. Very slow isolated notches may therefore produce wheel counts but no eligible reversal segment. A zero result from those movements is not proof that a fault is absent.
Reading reversals and uneven counts
The detector finds the majority direction within each eligible segment and counts events with the opposite sign. It does not know what your finger intended. Reversing your hand midway through a run can increase the count even when the mouse works correctly. A result matters when the same pattern appears during repeated, deliberately one-way movements.
Unequal upward and downward counts do not automatically indicate damage. You may have moved the wheel different distances, and software can combine or transform its input. Compare the symptom rather than demanding equal totals. The scroll test is useful for spotting repeated direction changes, but it cannot count physical encoder detents that never reach the browser.
Compare the mouse, the application and the connection
If the problem only occurs on one website, try a static document and another browser before changing hardware. If it follows the same mouse across applications, compare another connection or computer when available. Keep notes on whether the fault follows the mouse or remains with one software setup; that distinction gives the next check a clear purpose.
For a wider check of buttons and movement, use the mouse test. The how to test a mouse guide explains how to compare a suspect device with another input source. Complete those comparisons before buying a replacement based on a single unusual count.
Measurement limits
Wheel events are software observations, not a direct electrical measurement of the encoder. The operating system, driver and browser determine which deltas arrive. Pixel, line and page modes use different units, so a step of one in one mode is not necessarily smaller than a step of ten in another mode.
Trackpad inertia and free-spinning wheels can continue producing events after the initial movement. A browser test cannot reliably label those devices as mechanically faulty from event timing alone. It also cannot clean a wheel, repair an encoder or establish whether a warranty applies. Use repeatable evidence to describe the symptom when contacting support.
03
Mouse scroll test questions
How do I use this mouse scroll test to find a skipping wheel?
Scroll in one direction without rocking the wheel, pause, and repeat in the other direction. Compare the reversal count with what you physically did. A repeatable opposite-direction event is useful evidence, while a missed notch may leave no event to count. Check the same symptom in a second application.
Why does the scroll wheel test show a direction change I did not make?
The page received a delta opposing the majority direction of a short movement. That can accompany a wheel fault, but input processing and an accidental change of direction can produce a similar record. Repeat with a steady hand and compare another mouse or application before attributing the event to the encoder.
What does delta mode mean in a scroll wheel tester?
Delta mode identifies the units used for the wheel event: pixels, lines or pages. The displayed step size is meaningful only alongside that mode. It is not a calibrated physical distance and does not measure how far you turned the wheel. Compare runs made with the same device, mode and settings.
Can a mouse scroll test fix a wheel that jumps?
No. This page records input and helps you reproduce the symptom; it does not change the mouse firmware or repair moving parts. If the fault appears in one application only, investigate that application first. If it follows the mouse across computers, consult its support or repair instructions with your observations.
Does this scroll test work with trackpads and free spin wheels?
It can display their wheel events, but its reversal count needs more cautious interpretation. Inertial scrolling may continue after a gesture, and free spin does not correspond to discrete notches. Use consistent gestures, wait for movement to stop between directions, and avoid treating uneven event sizes as evidence of physical damage.