Typing Accuracy Test: measure precision, not speed
A typing test where accuracy is the score and speed is the footnote. You type a passage of real sentences, with capitals, commas and apostrophes, as carefully as you can. The result reports your accuracy, how many characters landed, how many did not, how many corrections you made, and which specific keys produced the errors. Free, no sign-up, and no time limit pressuring you into mistakes.
5 sentences, with capitals and punctuation. There is no clock. Type them as accurately as you can, and correct anything you notice, exactly as you would in real work.
Click the text to begin. Nothing is timing you. Accuracy is the score here.
Where to go from here
Why accuracy is measured before speed
Every keystroke you make is a repetition, and repetitions build habits whether or not they were correct. Practising at 88% accuracy is practising the wrong movement roughly one time in eight, and those repetitions do not cancel out: they compete with the correct ones for the same motor pattern. This is why the standard advice is to slow down until accuracy recovers, and why it feels so counterproductive while you are doing it.
The economics are also plainly in accuracy's favour. Every uncorrected error you go back for costs a backspace, a re-type and the pause where you noticed. A mistake is rarely worth one character; it is usually worth three or four, plus the interruption. That is why a careful typist at 45 WPM regularly finishes real work faster than a fast one at 60.
What the numbers on this page mean
Accuracy is the share of keystrokes that were correct on the first attempt. Backspacing to fix something does not restore the accuracy figure, because the mistake still happened; it only restores the text.
Corrections counts the backspaces you used. Read it against accuracy: high accuracy with many corrections means you are second-guessing text that was already right. Low accuracy with few corrections means errors are getting through into the finished text, which is the more expensive failure.
Speed is reported, quietly, at the bottom. It is there so you can see the trade you made, not because it is what this page is for. Expect it to be lower than your speed-test result. That is the test working.
Where accuracy usually breaks first
Almost nobody loses accuracy on the home row. It goes on the reaches: the top-row stretch to P and Q, the bottom-row drop to B and N, and above all anything that needs a shift. The little fingers do the most awkward work on a keyboard and get the least practice, which is why a capital letter at the start of a sentence is one of the most commonly fumbled keystrokes there is.
The second place it goes is at speed changes. A word you know well comes out as a single burst, and the word after it needs a reach you have to think about. The error usually lands in the gap, on the first character of the harder word, because the hand was still travelling at the pace of the easy one. This is why an even rhythm is worth more to accuracy than raw care is.
How to raise your accuracy
Slow down by about ten per cent, not by half. The goal is the slowest pace at which you still feel fluent, because dropping to a crawl teaches a different, equally unhelpful rhythm.
Stop looking at the keyboard, even though it will make things worse for a week. Accuracy that depends on visual checking has a hard ceiling, because your eyes cannot be on the screen and the keys at once, and every glance is a pause you are paying for.
Then work on the specific keys rather than on "accuracy" in general. The error report below your result names them. Three letters practised deliberately for a week will do more than a month of typing carefully in general and hoping.
Questions people ask
What is a good typing accuracy?
95% is the working threshold: below it, practice is reinforcing errors as much as correcting them. 98% and above is strong, and roughly matches the finished-text accuracy measured in large-scale typing research, where the average uncorrected error rate was 1.17%.
Should I fix my mistakes or type through them?
In a test, fix them: it reflects what you would do in real work, and the corrections themselves are a measurement worth having. In practice drills, it depends on the drill. What matters is being consistent, because a run where you ignored errors is not comparable with one where you fixed them.
Why is my accuracy lower here than on a speed test?
Because this passage has capitals, commas and apostrophes in it. A test made of lowercase common words never asks your little fingers to reach for a shifted key, which is exactly where most people's accuracy actually falls apart.
Does accuracy really matter more than speed?
For getting work done, yes, up to about 98%. Beyond that the returns flatten and speed becomes the thing worth working on. Below 95%, speed practice is close to wasted, because you are getting faster at making the same mistakes.
Further reading
A test measures. It does not teach.
KeyTopia turns the keys that slowed you down into the practice text for your next session, and keeps doing it as they improve. It is free, carries no advertising, and works without an account until you want your progress to follow you between devices.