AcademyQWERTY vs Dvorak vs Colemak vs Workman: Layouts Compared

The short version:

• QWERTY is the default nearly everywhere. In a study of 168,960 typists, 98.1% used it (Dhakal et al., 2018).
• Dvorak, Colemak, Colemak-DH and Workman put the most common letters under your resting fingers. In our model below, they roughly double the share of typing done on the home row and cut awkward same-finger sequences.
• What the evidence doesn’t show is that switching makes an experienced typist faster. A controlled government study from 1956 found retrained typists needed weeks just to get back to their old speed.

Why QWERTY looks the way it does

Christopher Latham Sholes and his colleagues patented an early typewriter in 1868, with an alphabetical, piano-like keyboard. The arrangement changed over the following years; an 1873 prototype was used to demonstrate the machine to Remington, which began selling a typewriter in 1874. When several of the largest typewriter makers merged into the Union Typewriter Company in 1893, QWERTY became the de facto standard (Smithsonian Magazine).

The popular story says the layout was designed to slow typists down so the typebars wouldn’t jam. That story is contested. In a 2011 paper, Kyoto University researchers Koichi and Motoko Yasuoka traced the keyboard’s development alongside its early users and argued that it evolved in response to telegraph operators transcribing Morse code, not to slow anyone down (as reported by Smithsonian Magazine).

QWERTY outlived the typewriter because millions of people had already learned it.

The alternatives

Dvorak (1936)

August Dvorak and William L. Dealey filed their patent for a “typewriter keyboard” in 1932, and it was granted in 1936 (US Patent 2,040,248). The Dvorak layout puts all five vowels on the left hand’s home row and common consonants such as D, H, T, N and S on the right’s, and in our model below it has the highest share of letter pairs typed with alternating hands. Windows includes a United States-Dvorak layout (Microsoft Learn).

Colemak (2006)

Created by Shai Coleman and released on January 1, 2006, Colemak moves 17 keys from their QWERTY positions, and common shortcuts such as Ctrl+Z, X, C and V stay where they were (Colemak). According to the project, Windows 11 version 24H2 and later include Colemak as a pre-installed layout (Colemak FAQ).

Colemak-DH (2014)

Colemak Mod-DH, launched in October 2014 by a designer known as SteveP, moves D and H, “the 8th and 10th most common letters in English,” out of the center columns, where standard Colemak puts them and where reaching them takes a sideways movement of the hand. No keys change fingers compared with Colemak, and there are versions for ANSI, ISO and matrix keyboards (Colemak Mod-DH).

Workman (2010)

Published by OJ Bucao on September 6, 2010, Workman aims to reduce sideways movement of the fingers and wrists, on the view that side-to-side movement is more strenuous than moving up and down, and to put frequent keys under stronger fingers. It changes 21 characters from QWERTY (Workman).

Race it now

60secondsGrand Prix1-minute typing test

Measure your speed and accuracy with the standard one-minute benchmark.

• 1-minute test
• 1,000 common words

Start Grand Prix

How they compare: a model

Each layout’s designers publish figures, but they use different texts and different measures, so we ran one model over all five. TypeTrance’s layout model counts where each letter of a large sample of text lands under each layout, using that layout’s standard finger assignment on a row-staggered keyboard.

How to read it:

• Home row: the share of letter keystrokes typed without leaving the home row.
• Same-finger pairs: the share of consecutive letters typed by the same finger on two different keys, so one finger has to move twice in a row.
• Hand alternation: the share of consecutive letters typed by different hands.
• Top-to-bottom row jumps: the share of consecutive letters that jump between the top and bottom rows, skipping over the home row.

Method. A model estimate computed by TypeTrance on October 29, 2026, over 200,000 words of our English 1,000-word test text, generated the same way our timed tests generate it. Letters only; spaces and Shift are ignored. Runs on our 200-word and 10,000-word lists moved individual figures by a few points but told the same story: QWERTY keeps about a third of letters on the home row, and the four alternatives keep about two-thirds or more.

These numbers describe where letters sit. They say nothing about the speed anyone will reach, they ignore the cost of learning, and they depend on the text: code, other languages or your own writing would shift them.

What the evidence on speed says

The case for Dvorak once rested on studies run in the 1940s for the U.S. Navy. Reviewing them in 1990, the economists S. J. Liebowitz and Stephen E. Margolis pointed out that Dvorak was reported to have conducted the experiments himself and had a financial stake in the keyboard. They also described a 1956 U.S. General Services Administration study by Earle Strong: ten government typists retrained on Dvorak needed well over 25 days of four-hour-a-day training just to catch up to their old QWERTY speed, and afterward progressed more slowly than QWERTY typists given further QWERTY training. Strong concluded that Dvorak retraining would never pay for itself (Liebowitz and Margolis, 1990).

That study covers experienced typists only, so it can’t tell us whether complete beginners learn faster on Dvorak. Smithsonian Magazine summarizes the wider record the same way: some research showed Dvorak users reporting better speed and accuracy, though much of that testing was funded by Dvorak, and other research has suggested Dvorak isn’t more efficient. Colemak and Workman are newer, and we are not aware of controlled studies of their effect on typing speed.

What switching costs

• Relearning time. Expect a long stretch below your current speed. In the 1956 study, experienced typists needed more than 25 days of four-hour training just to get back to where they started.
• Shortcuts and muscle memory. Copy, paste, undo and editor commands move with their letters, except where a layout preserves them; Colemak keeps Ctrl+Z, X, C and V in place.
• Other people’s computers. With nearly everyone on QWERTY, shared and work computers will usually be set up for it, so you will either change settings each time or keep both layouts in your hands.
• Your history. Your old results stop being a fair comparison while you relearn.

Who might switch

• You type a lot and want a layout that, by design, keeps more of your typing on the home row, and you accept a long, slow transition.
• You are learning to touch type from scratch anyway, understand the trade-offs above, and enjoy the experiment.
• You are curious and have time to practice.

Who probably shouldn’t: anyone who needs their current speed for work in the next few months, or who types on shared computers every day.

Try a layout before you commit

TypeTrance can emulate each of these layouts on the keyboard you already have: open Settings, then under Keyboard choose a layout in “Practise another layout on this keyboard”. Emulation maps the physical position of each key to the new layout, so your system settings stay as they are. Set “Layout you type in” to your real layout so the keyboard heatmap shows the right keys. A week of short sessions on the 1-minute test will tell you more about how a layout feels than any table.

If you decide to learn a new layout properly, the staged plan in touch typing applies to it just as it does to QWERTY.