How many hours a week to learn a new skill?
Two to five, split across at least two sessions, is the range where most adults make steady progress. The floor matters more than the ceiling: below about two hours a week you spend most of your time recovering ground you lost since last time, so the same annual total buys far less than it should.
Put your real weekly hours in and it prices a whole subject against them — 47 hours of SQL, for instance, in weeks rather than in encouragement.
See what your week actually buysThe question people mean is about total; the answer is about spacing
A hundred hours in a year is a hundred hours whichever way you cut it, and it is very much not the same skill at the end. Twenty-five four-hour Saturdays and a hundred one-hour evenings are different interventions, and the research on this is unusually clear about which direction the difference runs.
Distributed repetition beats massed repetition for long-term retention, and the effect is largest exactly where you care about it — at delays of 24 hours and longer.1 The same review adds a wrinkle worth knowing: the optimal gap scales with how long you need to remember something for. Learning for next week and learning for next year want different schedules, and only one of those is what you are doing here.
- One 4-hour block a week — the worst common shape. Fatigue in the second half, six days of decay after.
- Two 2-hour sessions — noticeably better for the same total.
- Four 1-hour sessions — better still for anything with judgement in it.
- Daily 20 minutes — excellent for retention, hard for anything that needs a run-up.
The floor, and why it is a floor rather than a preference
Below roughly two hours a week, split into at least two sittings, most adults are running a losing race against forgetting. The material from three weeks ago has decayed enough that each session starts with recovering it, and recovery crowds out the new work. The hours are real and the progress is not.
This is not a motivational claim, and it is not about discipline. It falls straight out of the shape of the forgetting curve: most of the loss on untouched material happens fast, and a gap wide enough to lose it is a gap you then have to pay for twice.2
The practical consequence is that *one hour a week is usually worse than it looks and two is usually enough*. If two is genuinely all you have, spend it as two ones rather than one two, and put the recovery to work by starting each session with retrieval rather than with review.
What a realistic week costs in calendar time
Once the floor is cleared, the arithmetic is honest and dull. Our SQL course is 47 hours across 26 skills; Python fundamentals is 30. Divide by the hours you actually keep — not the ones you intend — and you have the answer nobody selling a course wants to print.
- Five hours a week — a subject this size in about ten weeks.
- Three hours — about four months, which is the most common honest answer.
- Two hours — about six months, and it works if the two hours are split.
- Ten hours — about five weeks, and almost nobody sustains it past three.
Consistency beats intensity, and the reason is not moral
The advice to "be consistent" usually arrives with a whiff of virtue about it, which makes it easy to dismiss. The mechanism is more interesting than the sermon: a consistent schedule keeps the gaps short enough that retrieval is still possible, and retrieval is what does the work.3 An irregular schedule produces gaps that are sometimes too long, and the sessions on the far side of those gaps are spent re-reading.
Which is why the useful thing to fix is not the total. It is the worst gap in your month. Two hours a week with a three-week hole in it behaves like a much smaller commitment than the number suggests, and the fix costs nothing.
Sources, and what each is worth
Questions that come with this one
Is one hour a week enough to learn anything?
For a small, self-contained skill, yes. For a subject with dependencies — where week six needs week two — one hour a week means most sessions start by rebuilding what decayed. Two hours split in half is a much bigger jump than the numbers suggest.
Should I study every day?
You do not need to, and for most adults it is a schedule that breaks. Two to four sessions a week captures nearly all of the spacing benefit without requiring your life to cooperate every single day.
What about weekends only?
Workable if it is both days rather than one long Saturday. Two sittings a week is the thing that matters; whether they are Tuesday and Thursday or Saturday and Sunday matters much less.
Does the answer change for physical skills?
The direction holds — spacing beats massing for motor skills too — but the literature is clear that the size of the effect varies by task type. Treat the numbers here as being about the kind of learning this site is for.
Put this into practice
Reading about how to learn is the easy half. Tell us what you are actually trying to get good at, and we will build the path — the skills in the order they depend on each other, and marked work to prove you got there.
Build my path