Reading Time Calculator

How reading time is worked out

The calculation behind every reading time estimate is one division:

Reading time in minutes = word count divided by reading speed in words per minute

Take a 1,500 word article and the average silent reading rate for English non-fiction, 238 words per minute. The division gives 6.3025 minutes, and converting the fractional part into seconds gives 6 minutes and about 18 seconds. Change the speed to 260 words per minute, the average for fiction, and the same article takes 5 minutes 46 seconds.

That is the whole method. What makes the result useful, or misleading, is the speed figure you plug in, so it is worth knowing where those numbers come from and how far they can be pushed.

Notice how much the speed moves the answer. The same 1,500 word text lands between about 5 and a half minutes and 8 minutes depending only on the speed assumption, which is why a reading time without its speed is half a statement.

Where the reading speed numbers come from

The most useful source is a 2019 meta-analysis by Marc Brysbaert, published in the Journal of Memory and Language and available open access at doi.org/10.1016/j.jml.2019.104047. It pooled 190 studies with 18,573 participants and reported an average silent reading rate for adults in English of 238 words per minute for non-fiction and 260 for fiction. The oral reading rate came out at 183 words per minute.

The reason non-fiction comes out slower is word length. Longer words take longer to process, and non-fiction text carries heavier vocabulary than a novel does. The review also reported that most adult readers of non-fiction fall somewhere in the band of 175 to 300 words per minute, and fiction readers between 200 and 320.

That range is the honest headline. Any tool that quotes a single speed without a range is hiding half the story, because individual differences between readers are large and only partly explained. Rates are lower for children, for older adults, and for readers working in a second language. The same estimates are gathered in the education literature, for instance in the review collected on files.eric.ed.gov, which cites the 238 and 260 figures when it builds reading time models for textbooks. The full study record is indexed at OpenAlex if you want the citation trail.

The 300 words per minute figure that used to be repeated everywhere is higher than this review found, and the difference comes from how the older estimates were built. The 2019 analysis restricted itself to reading for comprehension or pleasure, and it excluded studies that asked participants to memorise material or read single words. Under those conditions, people read faster.

What counts as a word

The word count is the other half of the estimate, and it is not a fixed thing. A hyphenated compound counts as one word under most counting rules, so a state-of-the-art solution is one word, not four. A number written with a digit and a unit counts as one token. Punctuation attached to a word does not create a new one.

Counting is also sensitive to what you leave out. A heading, a table of contents, a list of references, and the words in an image caption all add to the raw count without adding to the reading load in a meaningful way. On a long document, dropping the reference list can move the estimate by several minutes.

Two conventions matter for the time you publish. First, images, tables, equations and code blocks take time to interpret, and a page of dense tables can take longer to work through than a page of prose even though it holds fewer words. Second, a heading read on its own costs a moment of attention out of proportion to its length. Both effects push the real time above the word-based estimate, which is why a published read time is usually rounded up and labelled as an estimate.

Silent reading, reading aloud and skimming

One speed number cannot cover three different jobs, and mixing them up is the most common error in published read times.

Silent reading for comprehension runs at 238 words per minute for non-fiction. Reading aloud runs at 183 words per minute, so a script of 1,500 words takes about 8 minutes 12 seconds to speak. Skimming for headings and first lines is far faster; a headline scan of 1,500 words takes well under a minute, since the reader is sampling a small share of the text.

For listening, the number that matters is speech rate. People listen to narrated audio at roughly 150 to 160 words per minute, which is slower than silent reading because the listener has no control over the pace. That is why an audiobook at the same word count takes longer than the silent read, and why speed controls on audiobook players matter. The audiobook speed calculator handles the playback side of that trade.

Deep reading, the slow kind with close attention and rereading, has no single rate, because the reader is doing more than decoding words. A reasonable approach is to use the silent reading figure as a floor and treat anything that needs rereading as extra time on top of it. For study material, the reading level calculator gives a sense of whether the text is pitched right for the reader, which changes the time far more than a few words per minute do.

Turning the estimate into a schedule

The most useful form of the number is time per 1,000 words, because it scales to any length without redoing the arithmetic. At 238 words per minute, each 1,000 words takes about 4 minutes 12 seconds. At 260 it is 3 minutes 51 seconds, and at 183 it is 5 minutes 28 seconds.

That unit makes planning concrete. A 2,000 word article is about 8 minutes 24 seconds to read, and a 4,000 word report is about 16 minutes 48 seconds. A 240 page novel at roughly 300 words a page is 72,000 words, which is about 5 hours 3 minutes of reading at 238 words per minute, spread over however many sittings you take.

If you are scheduling rather than reading, add a buffer. Readers skim, they interrupt, and they read on a phone in pieces. A published read time that matches the raw estimate invites complaints from people who skim, so rounding up by 10 to 20 percent, or rounding to the next half minute, produces a number that holds up better in practice.

For the neighbouring tools, the word counter gives you the raw count for anything you paste, the text to speech tool turns a script into audio when reading aloud is the goal, and the average time calculator averages a set of durations when you are comparing several tasks.

Read time labels on articles and videos

Publishers put a read time next to a headline for one reason: readers use it to decide whether to start. The label works when it is honest about the speed it assumes, and it backfires when a five minute estimate turns into a fifteen minute read. Most news sites compute the figure from the word count at a speed between 200 and 240 words per minute, then round to the nearest minute.

Video and audio need a different conversion, because the viewer cannot go faster without a speed control. At a typical narration pace of 150 words per minute, a 900 word script is a six minute video, and a 1,500 word script is ten minutes. That is why the same script that reads in six minutes on the page can run to ten minutes on screen.

Length also feeds search and engagement decisions. A piece that claims four minutes sets an expectation the first paragraph has to meet, and a piece that claims twenty minutes had better deliver something worth twenty minutes. The number is a promise about the reader's time, so padding it with words that do not earn attention is a costly trade.

One habit keeps the label useful: state the speed if the audience is likely to notice. A short note such as based on 230 words per minute takes no space and heads off the complaint that the estimate felt rushed.

Estimates, not promises

A reading time estimate is a planning number, and a few things sit outside it. The biggest is comprehension. A text that is read for understanding, checked against sources, and reread for detail takes far longer than the words divided by a rate can express, and for dense material the difference can be two or three times. The estimate works best for prose read once for general understanding.

The second is difficulty. Technical writing, unfamiliar notation, and a second language all slow a reader down without changing the word count. The 2019 review is clear that rates drop for readers working in a second language, and the band of 175 to 300 words per minute is wide enough that two readers of the same page can differ by a factor of nearly two.

The third is the task on the screen. A phone in a queue and a laptop at a desk are different reading environments, and interruptions cost more than the average rate suggests.

With those limits in mind, the number is still worth having. It tells a writer whether a piece fits a lunch break, tells a student whether a chapter is a sit-down job or a coffee-break job, and tells an editor whether a long post is worth splitting. Set the speed to match your material, state it alongside the time, and treat the result as a good approximation rather than a promise.