WPM stands for words per minute. In reading, it is the number of words in a text divided by the number of minutes you took to read it. Someone who reads a 1,190-word article in exactly five minutes has read at 238 WPM.

It is a rate, like kilometres per hour. It tells you how quickly the words went by, and nothing about how much of them you understood.

The formula, two ways

The definition uses minutes:

WPM = words ÷ minutes

A stopwatch gives you seconds, so the form you will actually use is:

WPM = words × 60 ÷ seconds

The two are the same calculation. Multiplying by 60 converts “words per second” into “words per minute”.

Worked example

You read a 364-word passage in 1 minute 31 seconds.

  1. Convert the time to seconds: 60 + 31 = 91 seconds.
  2. Multiply the words by 60: 364 × 60 = 21,840.
  3. Divide by the seconds: 21,840 ÷ 91 = 240 WPM.

Using minutes instead gives the same answer: 91 seconds is 91 ÷ 60 = 1.5167 minutes, and 364 ÷ 1.5167 = 240.

The mistake to avoid

A time of 1 minute 31 seconds is not 1.31 minutes. Seconds are sixtieths of a minute, not hundredths. If you divide 364 by 1.31 you get about 278 WPM, which overstates the true rate by 38 WPM. Converting everything to seconds first removes the problem.

Turning the formula around

Once you know a WPM figure, the same relationship answers two other questions.

  • How long will this take? minutes = words ÷ WPM. A 1,500-word document at 240 WPM takes 1,500 ÷ 240 = 6.25 minutes, which is 6 min 15 sec.
  • How much can I read in the time I have? words = WPM × minutes. Twenty minutes at 240 WPM covers 240 × 20 = 4,800 words.

Reading WPM is not typing WPM

The abbreviation is shared with typing tests, and the two numbers are often confused. They measure different things in different units.

Reading WPM Typing WPM
What is measured Intake: text you take in Output: text you produce
What a “word” is An actual word in the text By common convention, five keystrokes
What limits it Recognising words and building meaning Finger movement and accuracy
What accompanies it A comprehension score An accuracy or error rate

Typing tests usually standardise the word at five keystrokes so that scores are comparable whatever text is used. Under that convention, 250 keystrokes in a minute scores 50 WPM, no matter if that was forty long words or sixty short ones. Reading WPM does no such thing. It counts the real words on the page, so a text full of long words yields a lower figure for the same effort.

That is why the two numbers cannot be compared. A person who types at 60 WPM and reads at 240 WPM does not read “four times faster than they type” in any useful sense; the units are not the same.

Why “word” is fuzzier than it looks

The formula depends on a word count, and counting words involves choices. Is “well-known” one word or two? Is “3.5” a word? What about “don’t”, a web address, or two words joined by an em dash with no spaces?

Unicode Standard Annex #29, the technical standard that defines default word boundaries for text, is explicit that there is no single definition of a word that suits every language and purpose. Some writing systems, including Chinese, Japanese and Thai, do not put spaces between words at all, so software needs a dictionary or a segmentation algorithm to find the boundaries.

For text written with spaces between words, this site first treats em dashes, en dashes and double hyphens as separators, then splits on whitespace and counts each chunk containing at least one letter or digit. “Don’t” is one word, “well-known” is one word, and “3.5” is one word. It uses browser word segmentation for supported scripts written without spaces, with an approximate fallback described in the methodology. Another tool may make different choices and report a different total for the same text.

For ordinary English prose the disagreement is usually small. As an illustration, if one counter reports 2,000 words and another reports 2,040 for the same article, the reading time at 238 WPM moves from 8 min 24 sec to 8 min 34 sec. The practical lesson is to use the same counter for the text and for the calculation, and not to read meaning into a difference of a few WPM.

Word length matters more than counting rules. Brysbaert (2019) found that adults reading English average 238 WPM on non-fiction and 260 WPM on fiction, and attributes the gap largely to non-fiction using longer words. A WPM value always belongs to a kind of text. Our guide to average reading speed explains what those averages do and do not cover.

WPM without comprehension is half a measurement

You can make the number as large as you like by moving your eyes down the page faster. What you cannot do is keep understanding constant while you do it. Rayner et al. (2016) reviewed the evidence and concluded that speed and comprehension trade off: past a reader’s normal pace, going faster costs understanding.

A WPM figure is therefore incomplete until it is paired with a check that you understood the text. Compare two results on a four-question quiz:

Reader Raw WPM Questions correct Quiz score
A 320 2 of 4 50%
B 230 4 of 4 100%

Reader A has the higher rate and a weaker quiz result. Reader B’s attempt is a more cautious starting point for planning, but neither quiz score measures the percentage of the entire passage understood. Compare multiple unfamiliar passages before choosing a personal baseline.

The reading speed test on this site reports both numbers for this reason. It can also show a comprehension-adjusted score, which is raw WPM multiplied by the fraction of questions answered correctly: 320 × 0.5 = 160 for reader A and 230 × 1 = 230 for reader B. That adjusted score is a convenience defined by this site to keep the two numbers together. It is not a standardised or clinical measure, and it should not be compared with figures from other tests. The relationship itself is covered in our guide to reading speed and comprehension.

How WPM feeds the reading time calculator

A reading-time estimate is the formula turned around: minutes = words ÷ WPM. The reading time calculator counts the words in your text, divides by a WPM value and reports the exact time along with a rounded label, which is the exact minutes rounded up.

The WPM value is the only assumption in that calculation, so the choice matters. For 1,000 words:

  • 180 WPM gives 5 min 33 sec
  • 238 WPM gives 4 min 12 sec
  • 300 WPM gives 3 min 20 sec

If the estimate is for an audience, the 238 WPM benchmark is a reasonable default. If it is for you, a measured figure is better; the speed test can pass your result to the calculator as your tested speed. The step-by-step arithmetic, including the rounding rule, is in how reading time is calculated.

Getting your own WPM

You need three things: a text with a known word count, a timer and the formula above. The details that make the result trustworthy, such as using unfamiliar text and reading at your natural pace, are set out in how to measure your reading speed accurately.

Whatever figure you get, treat it as a description of one reading of one text. It is a useful planning number, and that is all it needs to be.