Accuracy Mode Typing Test
You cannot move past a mistake until you correct it. Errors cost time, not accuracy.
QWERTY layout assumed. Backspace corrects; uncorrected errors count against net WPM.
Sixty words, and you cannot move past a single one of them until you've typed it correctly — a wrong character doesn't advance the passage at all, so you notice it, retype it, and only then does the test let you continue. There's no way to finish this test with a mistake still sitting in your submitted text, because the test simply won't let you reach the end while one remains.
That's the opposite mechanism from most tests on the site, where a wrong character still moves you forward and only counts against you afterward, as an uncorrected error subtracted from your net WPM once the run ends — the mechanics of that formula are covered in full on How WPM Is Actually Calculated. Here, an error can never go uncorrected, because it's never allowed to stay in place long enough to become one.
With uncorrected errors removed from the equation entirely, the only thing left for a mistake to cost you is time — the seconds spent noticing it, undoing it, and typing the right character in its place, before the sixty-word target moves any closer to finished.
How This Test Works
The sixty words you're given are sized exactly to the target, the same way the Words Mode and 100-Word tests size theirs — nothing spare, nothing padded on. The rule that makes this test different has nothing to do with the passage itself and everything to do with what happens the moment you type something wrong.
When you type a character that doesn't match, the mistake is counted, but the caret doesn't move forward and the character isn't added to your finished text: you're simply asked, in effect, to try that position again. Because of this, your submitted sixty words are always fully correct once you finish — there's no such thing as an uncorrected error on this test, so the usual net-WPM error penalty never applies here the way it does everywhere else.
What does still move is the clock. Every rejected keystroke still takes real time to notice and replace, and your WPM is calculated from your total elapsed time — so a run full of corrected mistakes takes measurably longer to finish than a clean one, even though both submit an identical, error-free sixty words.
Who It's For
This suits typists who want their score to reflect only clean, correctly-typed output, with no ambiguity about whether an error was ever left in — useful for building the habit of noticing a mistake the instant it happens rather than several words later, which is a separate skill from raw speed.
It's a demanding, sometimes frustrating choice if you're already prone to overcorrecting small things that don't matter, since the test won't let you skip past a minor slip any more than a serious one — every mismatch, however small, gets treated identically. If a slower, error-tolerant pace under pressure sounds more useful to practice than a strict one, the Zero-Error Accuracy Drill offers a related but separately paced way to build the same habit.
Why There's No Such Thing as an Uncorrected Error Here
On a standard test, an uncorrected error is any character in your final submitted text that doesn't match the target — something that's only possible because the test lets you move past a mistake without fixing it. This test removes that possibility at the source: since the caret won't advance until the character is right, nothing wrong can ever make it into your finished sixty words. The concept the rest of the site calls an 'uncorrected error' simply doesn't apply here.
Where the Cost of a Mistake Actually Shows Up
With uncorrected errors off the table, a mistake's entire cost moves into elapsed time: the moment spent recognizing the character is wrong, the moment spent typing the correct one in its place, both added to your total before the sixty-word target is reached. A run scattered with small corrections will always finish slower than a clean run of the same sixty words, even if the finished text looks identical at the end.
A Fixed Target Makes the Comparison Fair
Because every attempt targets exactly sixty words, comparing your finishing time across attempts is a clean measure of improvement — unlike a timed test, where a faster typist simply produces more text rather than finishing sooner. Two attempts with very different numbers of corrections along the way still both end at the same sixty words, so the only variable left standing between them is time.
Frequently Asked Questions
Can I finish this test with any mistakes left in my final text?
No — the caret won't move past a character until it matches the target, so by the time you reach the end, all sixty words are necessarily correct.
Why doesn't my net WPM show an error penalty here the way it does elsewhere?
Because that penalty is only applied to errors left uncorrected in the final text, and this rule makes an uncorrected error impossible by design — every mistake gets fixed before you can move on.
Does this test still track how many mistakes I made?
Yes — each wrong keystroke is still counted as an error, even though it never appears in your finished text; it simply shows up as time lost rather than as a mark against your final accuracy.
Is this a good test for beginners still learning key locations?
It can be a slow, occasionally frustrating one for typists who are still hunting for keys, since every miss forces an immediate retry rather than letting you push through and fix things at the end — the Zero-Error Accuracy Drill offers a gentler, self-paced way to build the same habit first.