Typing speed statistics

The headline numbers

One caveat applies to every row from the 2018 study: the volunteers took an online typing test by choice, their average age was 24.5, three quarters were between 11 and 30, and 68% were in the United States.¹ People who seek out typing tests are probably faster than the population at large, so 52 WPM is best read as "what typing-test takers do," not "what everyone does."

Where the "40 WPM average" comes from

The most repeated typing statistic on the web is that the average person types about 40 words per minute. Its usual origin is a 1999 study by Karat and colleagues at IBM that measured 24 office workers entering text by keyboard and by speech recognition. Those participants averaged 32.5 WPM when copying text and 19.0 WPM when composing their own; grouped by skill, the fast typists averaged 40 WPM, the moderate 35, and the slow 23.³ The 40 figure is the fast group, not the mean, and the sample was two dozen people typing on 1990s hardware.

The 2018 Aalto and Cambridge dataset is roughly seven thousand times larger and puts the average of test takers at 51.6 WPM.¹ Both numbers are real; they measure different people doing different things. If you need one figure for "an average adult at a keyboard," 40 to 50 WPM is honest, and either end can be defended.

How speeds are distributed

The 2018 study reports percentiles rather than a neat bell curve; the distribution is skewed toward the fast end, with a long tail of very fast typists.¹

No primary source publishes a "top 1%" threshold. A normal-curve estimate from the study's mean and standard deviation lands near 99 WPM, but because the real distribution is skewed, treat that as a rough guide. The rankings on typing-game sites measure their own players, who are faster still, and are not population data.

Accuracy and error rates

Across 136 million keystrokes, 1.17% of typed characters were wrong and left uncorrected, and nine out of ten typists stayed under 2.66%. Corrections (backspace and retyping) made up 6.3% of all keystrokes, about 2.3 per sentence. Substitutions (the wrong letter) were the most common error at 1.65%, ahead of omissions (0.80%) and insertions (0.67%).¹

Speed and accuracy were not a trade-off in that data. Faster typists made fewer uncorrected errors and needed fewer keystrokes per character; the correlations were negative (r = -0.21 for uncorrected errors, r = -0.40 for keystrokes per character).¹ This is why keydrills holds a 90% accuracy bar before a lesson counts as passed and treats speed as a by-product of clean repetition.

The single strongest predictor of speed was rollover: pressing the next key before releasing the previous one. Rollover accounted for 25% of keystrokes on average and correlated with speed at r = 0.73; the fastest typists rolled over 40% to 70% of their keystrokes, the slowest almost never. Average time between keystrokes was 239 ms, about 120 ms for fast typists and over 480 ms for slow ones.¹²

Touch typing versus self-taught

In the 2018 sample, 72% had taken a typing course at some point. Trained typists averaged 54.4 WPM against 49.0 WPM for the untrained, a real but modest gap. Typists used 6.95 fingers on average; 47.6% used nine or ten. More fingers did correlate with speed (r = 0.38; the fast group averaged 8.4 fingers, the slow 5.3), yet the authors note that people not using all their fingers still reached speeds comparable to touch typists.¹

A closer look at 30 typists with motion capture found the same thing: touch typists averaged 57.8 WPM, non-touch typists 58.9 WPM, no significant difference, with 8.5 versus 6.2 fingers in use. What predicted speed was a consistent finger-to-key mapping, preparing the next keystroke early, and keeping the hands still.⁵ In a separate study of 48 typists, standard touch typists averaged 80 WPM to the self-taught typists' 72, and only the self-taught group slowed down when the keyboard was covered.⁶

The practical reading: you do not need all ten fingers to be fast, but you do need each key to belong to one finger, every time, without looking. Standard touch typing is simply the most reliable way to get there, which is why the keydrills lessons teach the conventional mapping and keep your eyes on the screen. Formal instruction is also rarer than people assume: among 1,301 French university students, 1.3% had ever had typing training.⁷

Age

The large 2018 keyboard study did not analyze speed by age or gender. The classic finding on physical keyboards is Salthouse's 1984 study of practising typists aged 19 to 72: typing speed did not decline with age, even though older typists had slower reaction times and tapping rates. They compensated by reading further ahead of their fingers.⁸

Mobile typing does show an age gradient, in a 2019 study of 37,370 volunteers:⁴

Women and men in that study typed at the same speed on phones (36.0 versus 36.1 WPM).⁴

Phones versus keyboards

People type about 15 WPM slower on a phone than on a keyboard: 36.2 WPM on average, with the top quarter above 44 WPM and a few above 80. Two thumbs beat one finger, 37.7 to 29.2 WPM, and 82% of people use two thumbs. Autocorrect users were faster (43.4 WPM); people who relied on word prediction were about 10 WPM slower. Uncorrected errors ran at 2.34%, twice the desktop rate.⁴

What jobs require

Court reporting speeds are on a stenotype, a chorded machine, not a QWERTY keyboard; the US median pay for court reporters was $72,420 in 2025.¹² We could not find an authoritative typing-speed requirement for general transcriptionists or for programmers; the figures circulating online for both come from job boards and blogs, so we leave them out.

Records

Guinness stopped listing computer-keyboard speed records after 1986 because tests on different machines and texts cannot be compared with any accuracy, which is why Blackburn's 212 WPM appears in Guinness editions from 1976 to 1986 and not after.¹⁴ Competition scores today come from timed races on short passages and are not comparable to a five-minute test of ordinary prose.

How long it takes to learn

There is no rigorous study that maps hours of practice to words per minute for adults. The best classroom evidence comes from 132 fifth graders who received daily keyboarding instruction for nine weeks: 107 of them reached the 20 WPM threshold the study treated as "touch typing," and students who finished below 20 WPM had roughly a coin-flip chance of keeping the skill two years later.¹⁸ A 2003 study in Human Factors found that trainees drift back to hunt-and-peck after a course unless practice forces their eyes onto the screen, because looking at the keys pays off in the short term even though it caps speed later.¹⁹

State curriculum guidance is the closest thing to a schedule: Wisconsin's Department of Public Instruction recommends 15 to 30 minutes a day for five to six weeks, with a target of five WPM per grade level.²⁰ Claims like "40 WPM in two weeks" are marketing, and personal reports of 100 WPM after 50 hours are anecdotes. The keydrills method plans for two to four weeks of short daily sessions for the same reason the research does: retention comes from daily repetition with your eyes up, not from long sessions.

Keyboard layouts

Alternative layouts are rarer than internet discussion suggests. In the 2018 sample, 98.1% of participants typed on QWERTY, so Dvorak, Colemak and everything else combined were under 2%.¹ No test site publishes layout market share. Whether Dvorak is faster has been argued since the 1950s and is not settled; the keydrills lessons support Dvorak and Colemak on the on-screen keyboard, but the evidence says the layout matters far less than consistent fingering and accuracy.

How keydrills measures speed

The typing test uses the same conventions as the research above: a word is five characters including spaces, net WPM counts only characters still correct at the end, accuracy is first-strike (a corrected typo still counts as a miss), and consistency measures how even your inter-key intervals were. Results stay in your browser. The full scoring rules are on the typing speed page.

Sources

1. Dhakal, Feit, Kristensson, Oulasvirta. Observations on Typing from 136 Million Keystrokes. CHI 2018 (Aalto University, University of Cambridge). userinterfaces.aalto.fi ↗
2. Aalto University press release: The traits of fast typists discovered by analysing 136 million keystrokes. 2018. aalto.fi ↗
3. Karat, Halverson, Horn, Karat. Patterns of entry and correction in large vocabulary continuous speech recognition systems. CHI 1999. dl.acm.org ↗
4. Palin, Feit, Kim, Kristensson, Oulasvirta. How do People Type on Mobile Devices? Observations from a Study with 37,000 Volunteers. MobileHCI 2019. userinterfaces.aalto.fi ↗
5. Feit, Weir, Oulasvirta. How We Type: Movement Strategies and Performance in Everyday Typing. CHI 2016. userinterfaces.aalto.fi ↗
6. Logan, Ulrich, Lindsey. Different (key)strokes for different folks: How standard and nonstandard typists balance Fitts' law and Hick's law. Journal of Experimental Psychology: Human Perception and Performance, 2016 (Vanderbilt University summary). news.vanderbilt.edu ↗
7. Pinet et al. Typing expertise in a large student population. Cognitive Research: Principles and Implications, 2022. link.springer.com ↗
8. Salthouse. Effects of age and skill in typing. Journal of Experimental Psychology: General, 1984. jimdavies.org ↗
9. US Office of Personnel Management. General Schedule Qualification Standards, Clerical and Administrative Support Positions. opm.gov ↗
10. National Court Reporters Association. Registered Professional Reporter certification requirements. ncra.org ↗
11. National Court Reporters Association. Registered Merit Reporter and Certified Realtime Reporter requirements. ncra.org ↗
12. US Bureau of Labor Statistics. Occupational Outlook Handbook: Court Reporters and Simultaneous Captioners. bls.gov ↗
13. Guinness World Records. Fastest typist on an electronic typewriter (Stella Pajunas, 1946). guinnessworldrecords.com ↗
14. Wikipedia. Barbara Blackburn (typist). en.wikipedia.org ↗
15. Guinness World Records. Fastest realtime court reporter, stenotype writing (Mark Kislingbury, 2004). guinnessworldrecords.com ↗
16. Journal of Court Reporting. Mark Kislingbury breaks own record. 2022. thejcr.com ↗
17. Das Keyboard. The fastest typist in the world wins the Ultimate Typing Championship. 2020. daskeyboard.com ↗
18. Ertl. Keyboarding instruction in elementary school: fifth-grade touch typing outcomes and retention. ERIC, 2007. files.eric.ed.gov ↗
19. Yechiam, Erev, Yehene, Gopher. Melioration and the transition from touch-typing training to everyday use. Human Factors, 2003. journals.sagepub.com ↗
20. Wisconsin Department of Public Instruction. Elementary keyboarding guidance. dpi.wi.gov ↗

Spotted an error or a newer study? Tell us on the feedback page.

Keep reading

• How to learn touch typing — Home row, fingering, accuracy-first practice, and a two-to-four-week schedule.
• How to type faster — Raise WPM by cleaning accuracy, rhythm, and weak keys — not by trying harder.
• What is a good typing speed — WPM explained, how keydrills scores a test, and realistic speed benchmarks.
• Home row and finger placement — Which finger owns which key, the F and J bumps, and the opposite-hand Shift rule.
• Typing.com and keybr alternatives — An honest comparison of keydrills, Typing.com, keybr, and Monkeytype.
• Best keyboards for learning to touch type — Seven honest picks and the four things that matter: layout, F and J bumps, key travel, wrist position.