What counts as a good typing speed?
Start with what nobody can tell you. There is no census of typing speed, no representative population survey, and no public dataset that would let anyone say "the 70th percentile is 63 WPM" and defend it. Every site that prints a crisp skill table is drawing lines that no measurement supports.
So the bands below are orientation, not data. They are deliberately approximate, and they describe what a speed feels like to work at rather than pretending to place you in a distribution.
| Rough band | What working at that speed is like |
|---|---|
| Below about 30 WPM | Typing is the slowest part of writing. Your hands wait on your eyes because you are still locating keys. |
| About 30 to 45 WPM | Serviceable. Email, forms and short notes are fine; long documents feel like work. |
| About 45 to 60 WPM | Typing roughly keeps pace with composing prose. Most people stop noticing the keyboard somewhere in here. |
| About 60 to 80 WPM | Above the 52 WPM average reported in the largest published typing study. Transcription and heavy writing stop being painful. |
| About 80 to 100 WPM | Fast by any reasonable standard, and faster than almost everyone measured in that study. |
| Above 100 WPM | Very fast. The fastest participants in that study reached around 120 WPM. |
Two warnings about tables like this one. First, the boundaries are conventions rather than findings — 59 WPM and 61 WPM are the same speed. Second, a WPM figure means nothing without the accuracy that produced it. On any test that reports net WPM, accuracy is already inside the number: only characters that are correct in your final text count, so 70 WPM at 99% accuracy and 70 WPM at 90% accuracy are not two grades of the same result — the second one required considerably faster hands to reach the same score.
What does the largest published typing study say?
The best-sourced modern figure comes from researchers at Aalto University and the University of Cambridge, who collected 136 million keystrokes from 168,000 volunteers in more than 200 countries and reported an average of 52 words per minute, with the fastest typists reaching around 120 WPM (Dhakal, Feit, Kristensson and Oulasvirta, "Observations on Typing from 136 Million Keystrokes", CHI 2018, DOI 10.1145/3173574.3174220).
The caveat is inseparable from the number, and any honest use of it repeats the caveat: participants were self-selected online volunteers. They went looking for a typing test. That is a population of people who care enough about typing to measure it, which is almost certainly faster than the population as a whole. "The largest published study of people who chose to take a typing test found an average of 52 WPM" is a defensible sentence. "The average person types 52 WPM" is not.
For dated historical context, the same researchers note that professional typists in the 1970s and 1980s were commonly in the 60 to 90 WPM range — a useful reminder that the ceiling has not moved much even as the machines got easier to press.
Why "good" depends on what you do all day
The same number can be plenty or nowhere near enough depending on the task, because the tasks are not the same task.
Transcribing is not composing. When you copy text that already exists, your hands are the only limit. When you write something new, you are thinking, and your fingers spend most of the time idle. One older measurement — from 1999, reported in Wikipedia's summary of typing-speed conventions — put average computer users at 32.5 WPM when copying text and 19.0 WPM when composing it. Treat the exact figures as historical, but the roughly two-to-one gap is the point: a typing test measures your transcription speed, which is the upper bound on your writing speed, not your writing speed.
Code is not prose. Symbols and digits sit away from the home row, capitalisation is load-bearing, and the five-character word convention gives a semicolon exactly the same weight as the letter E. Expect a lower WPM on code-shaped text than on plain English, and do not read that as a decline in skill.
Some jobs put a floor under it. Live chat support, transcription, dispatch and data entry are roles where typing speed is the throughput of the job rather than an incidental skill, and employers in those fields often set an explicit minimum. If you are typing toward a stated requirement, the requirement is the definition of "good" and this article is not.
How to read your own number
Four checks turn a score into information.
Check which metric you are looking at. Raw WPM counts every printable key you pressed, right or wrong. Net WPM counts only the characters correct in your final text. The two answer different questions, and a site quoting a headline speed without saying which one it means is not telling you much. Keyboardo reports both — see how typing speed is measured for the exact formulas.
Check the gap between them. If your raw WPM is far above your net WPM, errors rather than finger speed are what is holding the score down, and the fastest available improvement is to stop making them. That is not a motivational slogan, it is arithmetic: an uncorrected wrong character does not count toward net WPM at all, and correcting it spends time that does count against it.
Convert it to something concrete. A word is defined as five characters, spaces and punctuation included, so 60 WPM is exactly 300 correct characters in a minute, or five characters per second. Written that way the number stops being an abstract score. Read WPM vs CPM for the full conversion arithmetic.
Ignore any single test. One 15-second run is mostly noise: an easy word run, a lucky rhythm, a single stumble at second nine. Take the median of several runs at the same duration, on the same settings, and watch the trend. A Keyboardo profile shows the average of your last ten tests and your daily average across the past 90 days for exactly this reason. If you want a like-for-like comparison against other people, the leaderboard is restricted to a single fixed configuration — English 1k words, punctuation and numbers off — because otherwise the numbers would not be comparable at all.
Which "good typing speed" claims should you ignore?
Four of them, all repeated everywhere, none of them supportable.
"The average typing speed is 40 WPM." No traceable source exists for this figure. It is repeated across hundreds of typing sites, each apparently citing the others. The best-sourced number available is the 52 WPM above, and it needs its self-selection caveat attached every time. There is more on how this number came to be everywhere in what the average typing speed actually is.
Typing speed by age. The tables giving one figure for ages 6 to 11, another for teenagers and another for adults have no traceable origin. They look authoritative, which is exactly why they spread. No reliable public dataset breaks typing speed down by age, so Keyboardo does not publish one.
"Only N% of typists exceed 100 WPM." Any percentile claim needs a named dataset and a description of who was in it. Without those, the percentage is decoration.
"Barbara Blackburn holds the world record at 212 WPM." Not currently, and not that number. Guinness World Records removed electric- and computer-typewriter speed records from its 1987 edition onward, stating that records on such machines cannot be compared with any accuracy. Its listed typing records today are alphabet-typing records — the fastest time to type the alphabet is 3.25 seconds, set by Shashank in India on 25 February 2024. Guinness had historically listed Blackburn at 170 wpm peak and 150 wpm sustained; the 212 figure comes from a 1985 newspaper report of a brief burst.
The practical version of all of this: a good typing speed is a little above the one you have, typed clean, measured the same way each time. That target is always available, always honest, and never needs a statistic to justify it. Take a 60-second test and you have your baseline.