The arithmetic is trivial; the rate is not
Dividing words by words per minute takes a second. Choosing the divisor is where every reading-time estimate is won or lost, and the numbers in general circulation are mostly folklore. "The average person reads 200 words per minute" and "300 words per minute is average" are both widely repeated and neither traces to a systematic source.
The best current evidence comes from Marc Brysbaert's 2019 meta-analysis, which pooled 190 studies covering more than 17,000 participants. It reports a mean silent reading rate of about 238 words per minute for English non-fiction and about 260 for fiction, with reading aloud slower at about 183. The spread around those means is wide — the middle of the distribution runs roughly from 175 to 300 words per minute — which is why an estimate for a specific person should use that person's measured rate rather than a population average.
Rate also depends heavily on what you are reading and why. Fiction is faster than non-fiction because sentence structures repeat and vocabulary is familiar. A statistics textbook read for an exam is slower than a magazine feature by a factor of two or three, not because the reader changed but because the task did: you re-read, you stop to work an example, you go back a page. Legal and technical prose slows readers further still. If the text you are estimating is not ordinary connected prose, adjust the rate down before you trust the answer.
Turning pages into words without guessing
Page counts are the least reliable unit in publishing, because a page is a physical object whose word capacity depends on trim size, typeface, point size, leading and margins. A mass-market paperback and a large-format trade paperback of the same novel can differ by 40% in page count while containing exactly the same words.
The default here is 275 words per page, a reasonable middle for adult trade fiction and non-fiction, but you should replace it with a measured figure whenever the estimate matters. The method takes two minutes: count the words on three full pages chosen from different parts of the book, average them, and enter that. Count a full line, multiply by the number of lines on the page, and subtract for short lines at paragraph ends — that is accurate enough. For academic books set in 10 or 11 point with generous leading, 350 to 450 words per page is common; for children's chapter books, 150 to 200.
The calculator uses your words-per-page figure in two places: to convert pages into a word count when you are working from pages, and to report minutes per page in both modes. Minutes per page is the number most people actually plan with, because it converts "read chapter seven" into a slot in a calendar. At 275 words per page and 238 words per minute it is 1.16 minutes, so a 22-page chapter is a little over 25 minutes.
Front matter, endnotes, bibliographies and indexes are usually not read at prose rate and often not read at all. Enter the page count of the body text only, or your estimate will overshoot for any scholarly book, where notes can run to a fifth of the extent.
Worked example: a 320-page book in three weeks
You have been assigned a 320-page book and have 21 days. You count the words on three pages — 289, 302 and 276 — and take the average. You read non-fiction at a measured 210 words per minute.
- Average the sample pages. (289 + 302 + 276) ÷ 3 = 867 ÷ 3 = 289 words per page.
- Convert to a word count. 320 × 289 = 92,480 words.
- Divide by your rate. 92,480 ÷ 210 = 440.4 minutes.
- Convert to hours. 440.4 ÷ 60 = 7.34 hours of reading.
- Split across the deadline. 440.4 ÷ 21 = 21.0 minutes a day, and 320 ÷ 21 = 15.2 pages a day.
- Sanity-check the per-page figure. 289 ÷ 210 = 1.38 minutes per page, so a 30-page chapter is about 41 minutes. That matches the daily figure: 15.2 pages × 1.38 = 21.0 minutes.
Notice how much the words-per-page measurement mattered. Had you used the 275 default instead of the measured 289, the estimate would have come out at 419 minutes rather than 440 — a five per cent difference from two minutes of counting. Had you used a rate of 238 instead of your measured 210, the estimate would have been 389 minutes, and you would have run 51 minutes short over three weeks.
What the estimate does and does not include
The number this calculator returns is continuous reading time at a steady rate, and nothing else. Real reading sessions include settling in, losing the thread and going back, looking things up, and stopping. For planning a study schedule, add 20 to 30 per cent to the raw figure and you will be closer; for estimating the label on a blog post, the raw figure is what publishers use.
Reading for retention is a different task with a different rate. If you intend to remember the material — for an exam, a viva, a case — you will re-read passages, annotate, and stop to summarise. The literature on study reading consistently finds that effective study is slower and more repetitive than first-pass reading, so plan two passes at the calculated rate rather than one pass at half speed. That structure also makes the second pass much faster, because the first has already built the map.
Skimming is genuinely faster and genuinely worse. Rayner and colleagues, reviewing the evidence on speed reading, concluded that gains beyond about 600 words per minute come from sampling less of the text rather than processing it faster, and that comprehension falls accordingly. The two skimming rows in the calculator's table are there so you can see the trade in minutes, not as a recommendation.
If the daily workload the calculator returns looks impossible, the honest responses are to cut the reading list, to extend the deadline, or to read some of it at skimming rate on purpose and know which parts. Pretending a 400-page book fits into an evening is how reading gets abandoned. To check whether the text is as hard as it feels, run a sample through the Flesch-Kincaid grade level calculator; a grade level well above your comfort zone explains a slow rate better than any lack of discipline.
Reading time for common lengths at three rates
| Length | Approx. words | At 150 wpm | At 238 wpm | At 350 wpm |
|---|---|---|---|---|
| Blog post | 800 | 5.3 min | 3.4 min | 2.3 min |
| Long-form article | 3,000 | 20.0 min | 12.6 min | 8.6 min |
| Journal paper | 7,000 | 46.7 min | 29.4 min | 20.0 min |
| Undergraduate essay chapter | 10,000 | 1 h 6.7 min | 42.0 min | 28.6 min |
| Novella | 30,000 | 3 h 20 min | 2 h 6 min | 1 h 26 min |
| Short novel (200 pages) | 55,000 | 6 h 7 min | 3 h 51 min | 2 h 37 min |
| Standard novel (300 pages) | 82,500 | 9 h 10 min | 5 h 47 min | 3 h 56 min |
| Long non-fiction (450 pages) | 124,000 | 13 h 47 min | 8 h 41 min | 5 h 54 min |
Word counts for the page-based rows assume 275 words per page. Every time cell is words ÷ rate.
Assumptions that make an estimate wrong
- Using a population average as your own rate. Individual rates for fluent adults span roughly 175 to 300 words per minute. Time yourself on 500 words of the actual material and use that.
- Applying a prose rate to non-prose. Tables, equations, code listings and footnote-dense scholarship are not read at prose rate. Estimate them separately or accept that the total will be badly low.
- Trusting a publisher's page count. Page counts include front matter, notes, bibliography and index. Body text is often 15 to 25 per cent less than the printed extent of a scholarly book.
- Ignoring the difference between reading and studying. Material you must retain takes at least two passes. Budget accordingly rather than halving the rate, which understates the value of the second pass.
- Forgetting that fatigue is real. Rate falls over long sessions. A four-hour block does not deliver four times the pages of a one-hour block.
- Believing a speed-reading claim. Comprehension falls as rate rises above about 600 words per minute, because the eyes are sampling rather than reading. Treat high-rate figures as coverage estimates, not reading estimates.
Related planning and measurement tools
The single highest-value input to this calculator is your own measured reading rate, and measuring it takes five minutes with the reading speed calculator: time yourself on a passage of known length, count errors if you are reading aloud, and use the result here instead of the default. Teachers assessing children should measure with a one-minute oral probe rather than assuming an adult rate applies.
Going the other way, from an assignment written in pages to a word count you can plan with, is the job of the words to pages calculator, which accounts for font, size and line spacing rather than assuming a fixed words-per-page figure. If what you are timing is a talk rather than private reading, delivery rate is much slower than silent reading and the speech time calculator uses the right conventions.
Two readability tools help explain why one text takes far longer than another of the same length: the Flesch-Kincaid grade level calculator converts sentence length and syllable density to a school grade, and the Flesch Reading Ease calculator puts the same inputs on a 0-to-100 difficulty scale. Both take about a minute on a sample paragraph and both are better predictors of how slow a text will feel than page count ever is.