Words per Hour: How to Measure Your Writing Speed

How many words can you write in an hour? Divide a session’s new words by its minutes and multiply by 60. Copy-typing alone would reach about 2,730.
A man wearing a black wristwatch writes with a pen in an open notebook on a wooden table, with an open laptop and a smartphone behind it
Photo: Alejandro Escamilla on Unsplash

How many words you can write in an hour is a question the top-ranking guides answer with estimates. Ali Luke, whose post on Aliventures was last updated in January 2026, offers “a very rough rule of thumb” of one to two hours per 1,000 words. In her words, “many writers are looking at more like 500 words per hour,” while she herself can “normally draft 1,000 words of fiction in an hour” (Aliventures). River’s guide from 27 January 2026, signed by its co-founder Chandler Supple, works through a week of 3,500 words in seven hours and arrives at 500 an hour, and its FAQ stretches the range from 200 to 2,000 (River). Laurence MacNaughton’s often-linked post on doubling your speed gives no hourly figure at all (MacNaughton, 2018). None of the three cites a study.

That isn’t a failing of theirs, because nobody has measured an average for people composing their own prose. The only well-measured figure is for copying, where the average typist in the largest published study would manage about 2,730 English words an hour. Your own rate is something you have to time. Take the new words from a session, divide them by its minutes and multiply by 60. The sections below check the guides’ numbers against the research, look at two writers who timed themselves, and then set out a method and a calculator.

What the top guides say, checked

Luke’s post contains a worked example that holds up. You spend two hours on Monday, an hour and a half on Tuesday and another hour and a half on Wednesday, and you end with 3,000 words. “That’s a rate of 600 words per hour,” she writes, which means 1,000 words take you an hour and 40 minutes. Her typing figure is shakier. She says it takes “about 25 minutes to physically type 1,000 words,” based on “an average typing speed of around 40 words per minute,” a number she links to the typing-test site Ratatype. Typing tests count a word as any five characters, though, while real English words in the articles on this blog average 5.67 characters including the space. At 40 test words a minute, 1,000 real words take about 28 minutes. The largest typing study on record, which comes up again further down, puts the average at 51.56 test words a minute, and that would bring the time down to about 22 minutes.

River asks readers to track “every writing session” for “one full week” and then divide total words by total hours. That’s a sound start, and the guide adds the sensible warning to “use averages but understand your ranges.” Two of its rules rest on nothing the page cites. It states that “drafting speed is often 2-3x faster than revision speed,” and it tells you to set your daily goal at 70 percent of your calculated maximum. At 500 words an hour and two hours a day, that rule gives 700 words. It may well be a reasonable buffer, but it’s a rule of thumb, and the guide doesn’t say where it comes from.

MacNaughton’s post is about method more than numbers. He tells readers to “track your word output using a notebook or spreadsheet” and claims that “just tracking your writing speed will make it go up.” His evidence is his own books. The first, he writes, took “more than a year and a half,” the second nine months and the third four.

The oldest words-per-hour figure belongs to a Post Office man

Anthony Trollope spent 33 years at the British Post Office and wrote his novels around the job. In his autobiography, published in 1883, he describes sitting down at half past five every morning with his watch in front of him. He required of himself 250 words every quarter of an hour, and he reports that the words “have been forthcoming as regularly as my watch went” (An Autobiography, chapter 15). That’s 1,000 words an hour. He began each morning by spending half an hour reading over the previous day’s work, though, and he judged that “three hours a day will produce as much as a man ought to write.”

Trollope doesn’t explain in that passage how he counted, so his figure reads best as a target he held himself to. It still shows a professional novelist pacing himself against a clock.

A spreadsheet from 2011 shows how to measure yourself

The most detailed self-measurement I found is a blog post the fantasy author Rachel Aaron published on 8 June 2011 (Rachel Aaron, 2011). For two months she logged “the time I started, the time I stopped, how many words I wrote, and where I was writing.” The sheet showed a clear pattern. On days with one hour to write, she “never managed more than five hundred words in that hour.” On days with five hours of solid writing she was “clearing close to 1500 words an hour,” and the pace held until around the seventh hour, when she was, as she puts it, “simply too brain fried to go on.”

The change she credits most is small. Before each session she “dedicated five minutes (sometimes more, never less)” to writing out what was going to happen. Sometimes it wasn’t even a paragraph, she says, “just a list of this happens then this then this.” She calls them “these five stupid minutes” and reports that they took her from 2,000 to 5,000 words a day within a week, without adding to her five-hour block. She also found that “the afternoons I wrote at the coffee shop with no wireless were twice as productive as the mornings I wrote at home.” Place and time of day changed together there, so her sheet can’t say which of the two mattered.

Her rates belong to her, and so do the conditions she wrote under. What travels is the procedure, and the method below follows it.

How to measure your own speed

Aaron and River measure the same two things, the new words and the minutes of each session, and they differ mainly in how long they keep at it. River settles for “one full week.” Aaron kept her sheet for two months and noted the place as well, which is how she spotted the coffee shop. A week is enough for a first figure, and a longer log tells you more about the conditions behind it.

New words are the difference between the word count in your writing program before your first sentence and the count when you stop. A session in which you cut a lot can leave that number small or even negative, which marks it as revision, and revision belongs in a log of its own. Minutes run from the first sentence to the moment you stop. Short pauses to think stay in, since they’ll be there every time you write, while a phone call, lunch or twenty minutes in your inbox come out. Sessions under five minutes are best dropped altogether, because one sentence in two minutes would give you an hourly rate you’ll never see again.

Neither guide goes beyond a single average. Luke and River both divide all the words by all the hours, which gives the overall rate. The median is the middle value when the individual sessions are lined up by speed, and it describes an ordinary session better, because the overall rate leans towards the long ones.

An invented week shows how far apart the two can be, and the same numbers are preloaded in the calculator further down. On Monday evening the writer manages 340 words in 30 minutes, a rate of 680 an hour. Tuesday evening brings 560 words in 45 minutes, or 747, and Thursday evening 260 words in 25 minutes, or 624. The long Saturday morning produces 1,290 words in 90 minutes, a rate of 860, and Sunday evening closes the week with 170 words in 20 minutes, or 510. Altogether that’s 2,620 words in 210 minutes.

The overall rate is 2,620 words in three and a half hours, or 749 words an hour. The median session sits at 680. Saturday’s long morning produced almost half the week’s words on its own, and that’s what pulls the overall figure up. If your real weeks consist of short evening sessions, plan with the median, since those Saturday mornings won’t be there most of the time.

The calculator turns your sessions into planning numbers

The calculator below takes up to five sessions, each with its new words and minutes, and returns the overall rate, the median session and the range from slowest to fastest. From those it derives the time for 1,000 words and a word goal for a 25-minute sprint, and with a target length it adds the writing hours the project needs. The optional typing test field takes a result from any online test in words a minute and shows what share of that pure copying speed your writing reaches. The next section explains why the share comes out so small.

SessionNew wordsMinutes
1
2
3
4
5

With the preloaded week, the calculator gives 749 words an hour, a median of 680 and a range from 510 to 860. A thousand words take 1 hour 20 minutes at that pace, and the suggested goal for a 25-minute sprint is 312 words. An 80,000-word novel would need about 107 writing hours.

Why writing is so much slower than typing

The guides’ figures make more sense once they’re set against what has actually been measured, and typing has been measured on an enormous scale. In 2018 Vivek Dhakal and three colleagues published data from 168,960 volunteers who each typed 15 English sentences in an online test (Dhakal et al., CHI 2018). The instructions told them “to first read through and memorise a sentence, then type it as quickly and accurately as possible.” They averaged 51.56 words a minute, with a word defined as “any five characters,” and the paper notes that the fastest typists reach “speeds of even 120 WPM or more.” Converted to real English words, the average comes to roughly 2,730 an hour.

That number puts the top of River’s range in perspective. A writer composing 2,000 words an hour would be working at almost three quarters of the average copying speed, with barely any time left for deciding what to say. Luke’s 500 an hour, by contrast, is under a fifth of it, and the invented week above, at 749, is a little more than a quarter. The gap has a well-documented cause. In the 1970s John Gould, a researcher at IBM’s research centre in Yorktown Heights, New York, and his colleague Stephen Boies had people compose letters by writing, by dictating and by speaking them. Their abstract in Science is plain about the result. “Planning time is about two-thirds of composition time, regardless of the method of composition” (Gould and Boies, Science 1978).

MacNaughton’s experience with dictation fits the same picture. He found he could dictate “nearly twice as fast” as he could type, which is a big gain, yet far smaller than the raw speed of speech would suggest. Gould saw the same thing in a second paper, published in the Journal of Experimental Psychology, which followed eight highly experienced business executives. Going only by how fast each method lets words out, dictating should have been about 500 percent faster than writing by hand. The executives dictated 20 to 65 percent faster than they wrote (Gould, 1978). The abstract sums it up in a single sentence, “Output modality is not the limiting factor in composition.”

Phones show the pattern from the other side. A Stanford team led by Sherry Ruan had 48 people, half of them native English speakers and half native Mandarin speakers, transcribe short phrases by voice and by the on-screen keyboard (Ruan et al., 2018). The abstract reports that English input by speech “was 2.93 times faster,” and the paper’s tables give averages of 152.86 and 52.24 words a minute. Those participants were copying phrases they were shown, so nothing had to be planned. Set next to MacNaughton’s factor of two, the comparison suggests that dictation software helps most when you already know what you’re going to say.

That’s also what the share in the calculator is about. If planning takes about two-thirds of the time, roughly a third is left for getting words down, so a share somewhere near 30 percent is what you’d expect. That’s my own back-of-the-envelope reading of Gould’s figure, and his subjects wrote business letters rather than chapters. If your share comes out well below that, you’re doing a lot of your planning mid-session, and a touch-typing course won’t change much.

What your number does and doesn’t tell you

When I lined up the published figures, only one of them turned out to be a measurement in the strict sense. Trollope’s is a target he set himself, and Aaron’s comes from two months of her own logs. Luke’s and River’s are estimates, and MacNaughton writes that dictation came out “nearly twice as fast as I can type” for him, without putting an hourly figure on either. The typing rate is the only one with a large sample behind it.

Published writing speed figures and where they come from
SourceFigureWhat kind of figure
Anthony Trollope, autobiography, 18831,000 words an hour, three hours a daya target he set himself, reported afterwards
Rachel Aaron, 2011500, later over 1,500 words an hourher own session logs over two months
Ali Luke, Aliventuresabout 500 for many writers, 1,000 for herestimate
River guide, 2026500 in a worked example, 200 to 2,000 in its FAQexample and estimate
Laurence MacNaughton, 2018dictation nearly twice his typing speedhis own experience
Dhakal et al., 2018about 2,730 words an hour when copying, convertedmeasurement, 168,960 people

None of this gives you a scale on which 600 is slow and 1,200 is fast. If you write 400 words an hour, you’re starting a little below where Aaron started, and that’s all the comparison can tell you. Your figure becomes meaningful when you set it against your own figures from last month, from another time of day or from another project. A scene you’ve already planned and a chapter that needs references probably won’t run at the same speed, though I found nobody who has measured it, so it pays to keep one rate per project.

Handwriting needs a separate count. Most published handwriting speeds I came across come from copying tasks, which says little about composing. If you draft on paper, How Many Words Is a Handwritten Page? shows how to turn pages into a word count you can time.

Faster isn’t automatically better

A higher rate means more text per hour, and it says nothing about the text. A set of experiments at the University of Waterloo even suggests that slowing down can change what people write. Srdan Medimorec and Evan Risko had students type essays with both hands or with only one, which slowed them down considerably (Medimorec and Risko, 2016). Essays typed one-handed used more varied and less common words.

The study isn’t an argument for slow typing either. In the second experiment, where essays had to be finished in 25 minutes, the effect on vocabulary didn’t appear. In the argumentative essays of the third, one-handed writers repeated the same content words across sentences less often, and the authors count that as a likely plus too, since expert raters in earlier studies scored such essays higher. But the essays were only scored by text-analysis software, so the study can’t tell you which versions readers would have preferred. The university’s own news release quotes the authors saying they’re not suggesting students write their term papers with one hand (University of Waterloo). So treat your speed as a planning input that you don’t need to push at any cost.

The figure also stops making sense once you start revising. Cutting a 300-word paragraph to 200 is good work and a negative words-per-hour score at the same time. Hours or pages edited are better measures for that stage, and Words, Hours or Pages? goes through which one fits which phase of a project.

Putting your number to work

Your rate pays off most in planning. Divide the length of your project by it and you’ll know how many writing hours it takes. At 680 words an hour, an 80,000-word draft needs about 118 hours, which is close to three hours a week for ten months. The Writing Goal Calculator turns that into a daily target for a deadline and the days you can actually write. How to Write a Book While Working Full Time runs the same sum for several weekly schedules alongside a day job. Daily vs Weekly Writing Goals: What the Evidence Shows covers a Danish experiment in which 78 students with daily goals completed 690 tables in a week and 77 students with one weekly goal completed 521.

For sprints, divide your rate by 60 and multiply by the length of the sprint. At 680 words an hour, 25 minutes comes to about 280 words. Writing Sprints explains why a goal drawn from your own pace holds up better than a borrowed one. And How Many Words a Day Should You Write? works through daily targets for different situations.

Measure again after two or three months. If you’ve started planning scenes in advance, as Aaron did, or booked longer sessions, the second round will show whether that shows up in words an hour. If nothing moves, that’s useful too, because it means more text will have to come from more hours.

Is 1,000 words an hour fast?

It’s at the upper end of what writers report for first drafts. Trollope held himself to that pace, and Ali Luke gives it as her usual drafting speed for fiction, while she reckons many writers manage about half. Whether it’s fast for you depends on what you’re writing and how much you planned before you sat down.

How long does it take to write 1,000 words?

At 500 words an hour it takes two hours, at 750 about 80 minutes and at 1,000 an hour. Time a few of your own sessions and the calculator above gives you the figure for your pace. Research, outlining and editing come on top of that.

Will practice make me faster?

Probably a little, and less than most people hope. In Gould and Boies’s studies, people with years of dictation experience were only 0 to 25 percent faster than beginners, depending on how complex the letters were (Gould and Boies, 1978). The bigger jumps, like Aaron’s, came from changing how she prepared each session.

How many words can you type in an hour?

When typing short sentences they had just read and memorised, the 168,960 people in Dhakal’s study averaged the equivalent of about 2,730 English words an hour, and fast typists reach more than twice that. That’s a ceiling for copying, and composing your own text stays well below it.

Share