The question has one answer that is short and wrong, and three answers that are longer and true.

Learning touch typing is not a single event, so it does not have a single duration. The mechanics, meaning which finger owns which key and how to find home row without looking, take hours of practice rather than weeks. Fluency, meaning you can type real sentences without glancing down and without stopping to think, takes a few weeks of short daily sessions for most people. Passing the speed you had before you started takes anywhere from about three weeks to several months, and the deciding variable is not talent or age. It is how fast you already were. A slow hunt-and-peck typist can overtake their old speed inside a month. A fast hunt-and-peck typist is paying off a bigger debt and can still be below their old number two months in.

That last part is the piece most timelines leave out, and it is the piece that makes people quit in week two. Here is the whole shape of it, with the parts that are actually documented separated from the parts that are not.

What are the three phases of learning touch typing?

Every honest version of this timeline has three phases, because three genuinely different things are being learned in sequence: a map, a habit, and a speed.

Phase 1 is mechanics. You are learning a lookup table. Index fingers anchor on the F and J bumps, each finger owns a column, and every reach returns home. This is knowledge, and knowledge arrives fast.

Phase 2 is fluency. You know where the keys are, but your eyes still want to check. This phase is about deleting the glance, not adding speed, and it is where the work is.

Phase 3 is speed. The map is automatic and the eyes stay up, so what is left is transitions: bigrams, punctuation, shift keys, and the rhythm that connects them. This phase does not end. It just gets slower per hour invested.

The table below gives the shape of each phase. The spans assume roughly 15 to 20 minutes of deliberate practice a day, which is what the documented accounts in the next section actually logged. They are directional, not measurements, and the speed column is deliberately expressed against your own starting point rather than as a WPM number, because the week-by-week WPM curves that circulate for beginners are not traceable to any measurement.

Phase	What you are learning	Typical span at 15 to 20 min a day	Speed band	You are through it when
1. Mechanics	Finger-to-key map, home row return, F and J bumps	First few days to about two weeks	Well below your old speed. Slow enough that it feels like going backwards	You can name the finger for any letter without looking, and your hands return to home row on their own
2. Fluency	Typing without glancing down, recovering from an error without checking	Roughly weeks two to eight	Climbing steeply, closing on your old speed	You can type a paragraph of ordinary prose, eyes on the screen the whole time, at 95% accuracy or better
3. Speed	Bigrams, punctuation, shift keys, rhythm, endurance	Month two onward, no fixed end	Past your old speed, then improving more slowly each month	Your sustained score on a longer test, not just a lucky one-minute run, is above your old baseline

Read that middle column as a range that your own starting point moves. Phase 1 is short for nearly everyone. Phase 2 and Phase 3 stretch or compress depending on what you are replacing.

How many hours does basic competence take?

There is no controlled study that pins this down, and the confident hour counts that circulate do not trace back to one. What does exist is better than a made-up average: people who kept a log.

A freeCodeCamp writer documented learning to touch type at 29 and published his weekly numbers. Inside 14 hours and 44 minutes of practice spread across about a month, he went from a week-one average of 15 WPM to 22 in week two, and 39 in week three, at which point he was typing as fast as he had with his old method. By the end of the month he was averaging around 50 WPM with a best near 61 and accuracy around 95%.

So the useful answer to "how many hours" is roughly this: the finger map costs a handful of hours, and the first genuinely comfortable, eyes-up typing shows up somewhere inside the first fifteen hours of practice if those hours are spread out. What it does not buy you is your old speed, which is a separate bill.

Why do existing fast hunt-and-peckers take longer?

Because your current method is not nothing. It is a trained skill with years of repetition behind it, and the size of the debt scales with how good it already was.

An SAP LeanIX engineer wrote up switching to touch typing after decades of typing his own way, and he did the one thing almost nobody does: he took a proper baseline first. Ten 60-second tests before starting, averaging about 77 WPM at 94.5% accuracy. Then he practised roughly 20 minutes a day. At the two-month mark, after about 19 hours of systematic practice across 61 days, he was averaging 72 WPM at 93%. Two months of honest work, and he was still five words per minute below where he started.

Compare the two accounts and the pattern is hard to miss. The writer starting from a high-30s hunt-and-peck speed passed his old number in week three. The engineer starting from a high-70s hunt-and-peck speed had not passed his after two months. Same technique, same rough daily commitment, opposite scoreboards, because they were climbing out of different holes.

Two things follow from that:

Take a baseline before you change anything. Without it you cannot tell progress from noise, and you will judge week two against a number you half remember.
Judge Phase 2 on capability, not on WPM. "Can I type this paragraph without looking down" is the milestone. Your speed will be ugly for a while and that is the process working, not failing.

If you are still building the map itself, start with the home row and finger placement guide before you start timing anything.

What daily schedule beats marathon sessions?

Short and daily, and this is one of the few places in typing advice with a proper study behind it, on typing specifically.

In 1978, Alan Baddeley and D. Longman published The Influence of Length and Frequency of Training Session on the Rate of Learning to Type in Ergonomics. They trained postal workers on a new keyboard using different session lengths and frequencies, and the group on the shorter, more widely spaced schedule learned more per hour of training than the groups doing longer blocks. The long-session groups finished the course in fewer calendar days, which looks like the faster route, and still got less typing skill out of each hour they spent.

Practically, that means:

15 to 20 minutes a day, most days. Both of the documented accounts above sit right in that band, and neither person was doing anything heroic.
Stop when accuracy falls. A tired session teaches your hands the errors you make when tired.
Weight the first two or three weeks heavier. The engineer's log notes exactly that, more practice up front while the map is still new.
Do not test constantly. Testing is measurement. Once a day at the end of a session is plenty, once a week is enough to see a trend.

The 30-day practice plan lays out what to put inside those 15 minutes once your fingers know the map.

When do you overtake your old speed?

Take your current speed and be honest about it, then use these markers.

If you are starting from a low baseline, under roughly 40 WPM, expect to match it in weeks rather than months and to pass it comfortably in the first couple of months. If you are starting from a genuinely fast hunt-and-peck speed, in the 70s or above, plan for months, not weeks, and plan to spend part of that time slower than you were.

For a reference line, the University of Cambridge summary of the largest typing study published reports that the average participant, out of 168,000 volunteers producing 136 million keystrokes, typed 52 words per minute, and the fastest reached 120 on randomised phrases nobody could rehearse. So a touch typist arriving in the 50s is at the average of everyone who takes a typing test, and the ceiling is a long way further up. What counts as a good typing speed puts the rest of the range in context.

The other thing worth knowing about Phase 3 is what actually separates fast typists from average ones. The same study found it is not finger count. It is overlap: the fastest typists press the next key before releasing the previous one, on 40 to 70 percent of keystrokes. That overlap only becomes available once your eyes are off your hands, which is the real reason the awkward Phase 2 weeks are worth sitting through.

How do you measure progress without fooling yourself?

Learning curves are noisy, and a noisy curve plus a hopeful learner produces bad conclusions in both directions. Four rules keep the numbers honest.

Fix the test length and never change it. A one-minute score and a five-minute score are different measurements of different things, and comparing across them will invent progress that is not there. Here is why your WPM differs on every typing test.
Take the median of three, not the best of five. Your best run is your ceiling on a good day. The middle run is your speed.
Read accuracy next to speed, always. A WPM gain paid for with accuracy is not a gain. In the engineer's log, speed and accuracy fell together, which is what a technique change in progress looks like.
Log every result and look at the line, not the last number. Week to week, the trend is the only thing that means anything.

TypeTimer is built for exactly that loop. Take the 1 minute typing test three times, keep the middle score, and every result is saved in your browser automatically. The progress panel then shows your personal best, how many tests you have taken, your last-10 average, your latest run, and a net WPM trend chart across your recent tests. Nothing is uploaded and there is no account, so the history is yours and stays on the device. When you want to check whether a gain survives fatigue, run the same protocol on the 5 minute test once a month.

Take your baseline today, before you change a single finger. It is the one measurement you cannot go back and collect later, and it is the number that turns the next two months from a guess into a graph.

Frequently asked questions

Can I learn touch typing in a week?

You can learn the map in a week. You will not be fluent in a week, and you will almost certainly be slower at the end of it than you were at the start. A week of 15-minute sessions is enough to know which finger owns which key and to return to home row without looking, which is real progress. Treat it as finishing Phase 1, not finishing.

Will my speed drop while learning?

Yes, and it is supposed to. You are removing a shortcut your brain relied on, which is looking at your hands, and the replacement is not built yet. How far it drops and how long it stays down depends on your starting speed. The documented case of a 77 WPM hunt-and-peck typist showed him still five WPM below baseline after two months of practice, while a high-30s typist matched his old speed in three weeks. Expect the dip. Judge yourself on whether you can keep your eyes up, not on the number.

Is 15 minutes a day enough?

Yes, and it beats the same total time in weekend marathons. That is the finding in Baddeley and Longman's 1978 typing study: shorter, more widely spaced sessions produced more skill per hour of training than longer blocks. Fifteen focused minutes daily, ending with one recorded test, is a schedule people actually keep, which matters more than any schedule they abandon.

Am I too old to learn touch typing?

No. The published first-person accounts of adults retraining are people in their late twenties, thirties and beyond, and they got there in weeks to months of ordinary daily practice. Age is not the variable that moves this timeline. Your existing typing speed is, and it works the opposite way to how people expect: the faster your current self-taught method, the longer the transition takes, because there is more habit to overwrite.

Sources

Dhakal, Feit, Kristensson & Oulasvirta, Observations on Typing from 136 Million Keystrokes, CHI 2018, Aalto University and University of Cambridge. 168,000 volunteers, 136 million keystrokes. Summarised by the University of Cambridge: 52 WPM average, 120 WPM fastest, 40 to 70 percent rollover among the fastest typists.
Baddeley & Longman, The Influence of Length and Frequency of Training Session on the Rate of Learning to Type, Ergonomics, 1978, vol. 21, pp. 627 to 635. Postal workers trained on a new keyboard under different session schedules.
I learned to touch type at the ripe old age of 29, freeCodeCamp. Weekly log: 14 hours 44 minutes of practice in about a month, 15 WPM in week one to roughly 50 WPM average at month end.
How to learn touch typing after decades of typing oddly, SAP LeanIX Engineering. Baseline of ten 60-second tests averaging 76.98 WPM at 94.5% accuracy, then 19 hours of practice across 61 days to 72 WPM at 93%.

How Long Does It Take to Learn Touch Typing?