Does typing speed matter in coding?

For most coding, no. Typing speed is rarely what limits a developer's day. Reading and understanding code takes up most of the time ¹, and faster typing does not shorten that reading. Speed matters more for the writing around the code, and touch typing helps mainly by letting you keep your eyes on the screen.

Reading code takes most of a developer's day, so typing speed has little room to help

In a field study of professional developers, the developers spent about 58 percent of their time on program comprehension activities ¹. Comprehension here means working out what code does, which is reading, tracing and reasoning. Typing speed only changes how long the writing takes, and the study's summary does not break the remaining time down into typing and other work.

We did not find a controlled study that measures how typing speed changes a programmer's output. The evidence on this question is indirect: where the time goes, and how fast people type in tests that are not about code. Read the rest of this page with that limit in mind.

Typing speed matters more for the prose around the code

Developers do write a lot of prose. In a 2018 NC State survey, programmers reported spending 26 percent of their work week writing, the lowest share of any field ². Those are self-reported figures, and they cover the whole work week, not only time spent on code.

That prose is where speed counts. Comments, commit messages, tickets, documentation and prompts to an AI coding tool are all typed or dictated text. If you write a lot of it, a steady and accurate rate saves real time. If you mostly edit code, it saves less.

Touch typing did not raise speed in the study that measured it

One direct test of touch typing is a 2016 transcription study. Touch typists entered text at an average of 57.8 words per minute ³, while typists who had never learned touch typing reached 58.93 words per minute in the same task ⁴. The untrained typists were slightly faster.

The study also found that self-taught typists, who never learned touch typing, reached speeds comparable with touch typists ⁵. So touch typing is not a requirement for fast typing. The test copied text, rather than writing code, so it says nothing direct about coding.

Touch typing does cut how often you look at the keyboard

The same study measured where people looked. Touch typists spent about 20 percent of their time copying text looking at the keyboard ⁶. Typists without touch-typing training spent about 40 percent ⁷.

This is the real benefit of touch typing in the study. When you glance down less, your eyes stay on the code. Whether that saves time during coding is plausible, but this study did not measure it during programming, so treat it as a likely benefit rather than a proven one.

In a 2006 study, expert Java developers found speech slower than typing

Speaking code is a different question from typing it. In a 2006 evaluation of a speech-based Java programming system, most expert Java developers had little trouble learning the system but found it slower than typing ⁸. That system is old, and newer tools differ, but the result warns against assuming voice is faster for code. For how voice coding tools work, see the guide to voice coding.

It depends on what your work is made of

What to do: measure your own speed on your own work

1. Time yourself on a paragraph of ordinary prose, and then on a block of comments or a commit message you would really write this week.
2. Note how much of your coding session went to reading and thinking, and how much went to typing. Your own session will tell you more than an average does.
3. If prose is the slow part, try dictation for long text before you spend weeks on drills. Is dictation faster than typing? covers when it is and when it is not.
4. If you are learning touch typing to look at the screen less, measure the change in your own sessions, not in a typing test alone.

For the general answer outside coding, read does typing speed still matter?. For a speed reference, see our entry on words per minute. If you are weighing voice for code, see how Talon voice coding works and whether Serenade still works for voice coding.