What WPM means

WPM is words per minute: a count of words divided by the minutes taken to read them. It is the unit both tools share. The reading time calculator takes a WPM and produces a time. The reading speed test takes a time and produces a WPM.

WPM depends on what counts as a word, so the two tools use one shared counter. A passage in the test has the same word count the calculator would report if you pasted it in.

How word count works

The counter uses the following explicit rules. Other tools may count some cases differently:

  1. Em dashes, en dashes and double hyphens are treated as separators, because writers often set them without spaces.
  2. The text is split on Unicode whitespace: spaces, tabs, line breaks and non-breaking spaces. Repeated whitespace makes no difference.
  3. Each remaining chunk counts as one word if it contains at least one letter or digit in any script. Chunks made only of punctuation, symbols or emoji count as zero.
Counting examples
Text Words Why
don’t stop 2 A contraction is one word, with a straight or curly apostrophe.
a well-known author 3 A hyphenated compound is one word.
fast—really fast 3 An em dash separates words even without spaces.
pages 10–20 3 An en dash separates the two numbers.
1,000 words in 3.5 minutes 5 Numbers count as words; internal commas and points do not split them.
Great job 🎉 !!! 2 Emoji and stand-alone punctuation are not words.

Some writing systems, including Chinese, Japanese and Thai, do not put spaces between words. For text in those scripts the counter uses the browser’s Intl.Segmenter word segmentation where it is available, and otherwise falls back to counting characters in those scripts, which is only an approximation.

There is no single correct definition of a word. The Unicode Consortium’s text segmentation standard makes the same point: word boundaries vary by language and by purpose. Other tools may count a hyphenated compound as two words or handle numbers differently, so a difference of a percent or two between counters is normal and rarely changes a reading-time estimate by more than a few seconds. Texts above 5,000,000 words are outside the range the tool accepts.

The secondary character count is the number of Unicode code points, including spaces.

How reading time is calculated

minutes = word count ÷ words per minute

For 1,000 words at 238 WPM: 1,000÷ 238 = 4.2017 minutes, or 252.1 seconds, shown as 4 min 12 sec.

Calculations use full floating-point precision. Rounding happens only when a result is displayed:

  • Under one minute: seconds, for example “42 sec”.
  • Under one hour: minutes and seconds to the nearest second, for example “4 min 12 sec”.
  • One hour or more: hours and minutes to the nearest minute, for example “1 hr 14 min”.

The rounded “min read” label

The label is the exact time in minutes rounded up to the next whole number, with a minimum of 1 for any non-empty text. 4 min 12 sec becomes “5 min read”.

Rounding up is deliberate. A label is a promise about the least time a reader should set aside. Rounding to the nearest minute would label a piece that takes 4 min 20 sec as a 4-minute read, which is less time than it needs. If you publish labels, use one rule on every page.

The reverse calculation

word budget = minutes × words per minute

Ten minutes at 238 WPM gives 10 × 238 = 2,380 words.

Reference values

Reading time for common word counts at the 238 WPM benchmark
Word countReading time at 238 WPMRounded label
10025 sec1 min read
2501 min 3 sec2 min read
5002 min 6 sec3 min read
7503 min 9 sec4 min read
1,0004 min 12 sec5 min read
1,5006 min 18 sec7 min read
2,0008 min 24 sec9 min read
2,50010 min 30 sec11 min read
3,00012 min 36 sec13 min read
5,00021 min 1 sec22 min read
10,00042 min 1 sec43 min read

Where the reading speeds come from

The default of 238 WPM is taken from Brysbaert (2019), a meta-analysis of 190 studies with 18,573 participants. Its main findings for adults reading English:

Average reading rates reported by Brysbaert (2019)
Kind of reading Average Typical range
Silent, non-fiction 238 WPM about 175–300 WPM
Silent, fiction 260 WPM about 200–320 WPM
Aloud 183 WPM —

The paper attributes most of the gap between fiction and non-fiction to word length: non-fiction uses longer words, so fewer of them fit into a minute of reading. It also notes that the frequently quoted figure of 300 WPM is higher than the evidence supports.

The calculator’s presets are:

  • Careful, 180 WPM. A round planning value near one end of the typical range. It is a convenience, not a research average.
  • Research benchmark, 238 WPM. The published non-fiction average described above.
  • Brisk, 300 WPM. A round planning value near one end of the typical range. It is a convenience, not a research average.
  • Custom. Any value from 50 to 1,500 WPM. Values outside that range are rejected instead of being silently adjusted.

These figures describe English. Reading rates measured in words per minute differ between languages, partly because languages package meaning into words of different lengths. Standardised test passages developed in 17 languages (Trauzettel-Klosinski & Dietz (2012)) illustrate how much the numbers depend on the language and on the text used.

Why results are estimates

The arithmetic is exact, but one of its inputs is an assumption. The calculator cannot know who will read your text or why. An average hides a wide spread between people, and each person’s speed shifts with the material. Treat the result as a reasonable central estimate, and choose a slower speed when being late matters more than being early.

Why text complexity matters

Eye-movement research shows where reading time goes. Readers do not sweep smoothly along a line. They make short jumps and pause on most words, and the length of each pause reflects how much work the word requires. Just & Carpenter (1980) found longer pauses on uncommon words and at points where a reader has to pull information together. Rayner (1998) reviewed two decades of this work: as text becomes harder, pauses lengthen, jumps shorten and readers look back at earlier words more often.

A word count sees none of that. Two 1,000-word texts can differ substantially in reading time if one is a familiar story and the other is a contract. Purpose matters as well: Carver (1992) described distinct reading processes, from scanning to memorising, each with its own typical rate. The practical remedy is in your hands: lower the speed for demanding material. See why difficult text takes longer to read.

Silent reading vs reading aloud

Reading aloud is slower than silent reading because speech has a physical speed limit. The calculator shows a secondary “read aloud” figure at 183 WPM, the adult oral reading average from Brysbaert (2019). It is labelled as rough for a reason: that average describes people reading a text aloud, not delivering a speech. Presentations, recordings and lessons are often slower, with pauses and emphasis. For anything spoken, enter a custom speed and confirm it by rehearsing against a clock.

How the reading speed test calculates WPM

WPM = passage word count × 60 ÷ elapsed seconds

  • Timing uses the browser’s high-resolution clock (performance.now()), which is not affected by changes to the system time.
  • The timer starts only when you press Start reading and stops when you press Finished reading. Repeated clicks on either button are ignored.
  • The passage is hidden while you answer, and question time is not counted.
  • Reloading the page cancels the attempt. Nothing is restored, so a partial attempt can never produce a score.

Attempts are checked for plausibility before any score is shown:

How unusual attempts are handled
Situation What happens
Finished in under 3 seconds, or faster than 1,500 WPM Not scored. You are asked to try again.
Timer ran longer than 30 minutes Not scored. The attempt is treated as interrupted.
Faster than 600 WPM Scored, with a note that the pace is more typical of skimming.
Slower than 60 WPM Scored, with a note that the attempt may have been interrupted.
You switched to another tab while reading Scored, with a note that the timer kept running.

Each passage was written for this site. None is taken from a book, news article or another website.

Why comprehension matters

A speed figure is only meaningful if the text was understood. Rayner et al. (2016) reviewed the evidence on speed reading and concluded that speed and comprehension trade off: people can go faster by skimming, but they take in less, and no training method has been shown to remove that cost. One reason is that understanding depends on being able to look back. Schotter, Tran & Rayner (2014) showed that comprehension falls when readers are prevented from re-reading earlier words.

The test therefore reports comprehension next to speed:

  • Comprehension score: correct answers out of four, and the same value as a percentage.
  • Comprehension-adjusted score: raw WPM × fraction correct. For example, 300 WPM × 0.75 = 225.

The adjusted score is a convenience metric defined by this website. It is not a standardised, clinical or educational measure, and it should not be compared with scores from other tests. When fewer than three answers are correct, the result carries a neutral note that a high WPM means less when comprehension is low.

Results are compared with the published average that fits the passage: 260 WPM for the narrative passage and 238 WPM for the others. The site shows no percentile rankings, because it has no population data to base them on.

What is stored on your device

  • Your tested reading speed, as a single number, if you choose to use it in the calculator.
  • Your last 5 test results: date, passage type, raw WPM, comprehension percentage and adjusted score.
  • Your reading-speed selection in the calculator.

Pasted text, passage text and individual answers are never stored or transmitted. Both tools work normally if local storage is blocked. Details are on the privacy page.

Known limitations

  • Benchmarks are English-language averages and do not transfer directly to other languages.
  • Averages describe groups. They do not predict any one reader.
  • Images, tables, code, formulas and footnotes are not accounted for by a word count.
  • The speed test uses short passages and four multiple-choice questions, which is a coarse measure that can be partly guessed.
  • Timing relies on manual button presses and includes a small reaction-time error.
  • Neither tool is a medical, psychological or educational assessment. They cannot identify dyslexia, ADHD or any other condition, and WPM is not a measure of intelligence.

References

  1. Brysbaert, M. (2019). How many words do we read per minute? A review and meta-analysis of reading rate. Journal of Memory and Language, 109, 104047. https://doi.org/10.1016/j.jml.2019.104047
  2. Rayner, K., Schotter, E. R., Masson, M. E. J., Potter, M. C., & Treiman, R. (2016). So much to read, so little time: How do we read, and can speed reading help?. Psychological Science in the Public Interest, 17(1), 4–34. https://doi.org/10.1177/1529100615623267
  3. Rayner, K. (1998). Eye movements in reading and information processing: 20 years of research. Psychological Bulletin, 124(3), 372–422. https://doi.org/10.1037/0033-2909.124.3.372
  4. Just, M. A., & Carpenter, P. A. (1980). A theory of reading: From eye fixations to comprehension. Psychological Review, 87(4), 329–354. https://doi.org/10.1037/0033-295X.87.4.329
  5. Schotter, E. R., Tran, R., & Rayner, K. (2014). Don't believe what you read (only once): Comprehension is supported by regressions during reading. Psychological Science, 25(6), 1218–1226. https://doi.org/10.1177/0956797614531148
  6. Carver, R. P. (1992). Reading rate: Theory, research, and practical implications. Journal of Reading, 36(2), 84–95.
  7. Trauzettel-Klosinski, S., & Dietz, K. (2012). Standardized assessment of reading performance: The new International Reading Speed Texts IReST. Investigative Ophthalmology & Visual Science, 53(9), 5452–5461. https://doi.org/10.1167/iovs.11-8284
  8. The Unicode Consortium (n.d.). Unicode Standard Annex #29: Unicode text segmentation. Unicode Consortium. https://unicode.org/reports/tr29/
  9. MDN contributors (n.d.). Intl.Segmenter. MDN Web Docs. https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Intl/Segmenter

If you find an error on this page, please let us know. Corrections are handled as described in the editorial policy.