How long should you study before a break?

What it costs on the clock

One full cycle is 2h 10m. The break after the last round isn’t counted — you’re done.

The number you were given, and where it came from

The 25 minute figure comes from Francesco Cirillo, who was a university student in the late 1980s and used a tomato-shaped kitchen timer. He has said plainly that he picked a short interval because it was short enough to start without arguing with himself, not because 25 minutes is a biological unit. It was a starting device, and it got frozen into a rule by everyone who repeated it afterwards.

The 90 minute figure comes from a different place: research on ultradian rhythms, cycles of alertness that run roughly 90 to 120 minutes through the day. That work was mostly done on sleep stages and then extended to waking attention, which is a bigger leap than most articles admit. It is a reasonable upper bound to think with, not a measurement of you.

The 52 minute figure comes from DeskTime, a time-tracking company, which looked at its own users' logs and reported what the most productive ten percent happened to do. That is an observation of a group that was already getting a lot done, not a trial where one group worked 52/17 and another did not.

What actually ends a study block

In practice, blocks end for three different reasons, and they want different fixes. The first is that you stopped understanding — you are reading the same paragraph again. The second is that you stopped caring — the work is still legible but you are checking your phone between sentences. The third is that your body objected: you are hungry, your neck hurts, you need the bathroom.

Only the first one is a real attention limit. The second is usually a sign the task is too vague to continue (you do not know what the next concrete step is), and the third is logistics. If you treat all three as 'I need a break', you will take breaks you did not need and skip the one you did.

A way to find your own number in three days

Pick a fixed block length — 30 minutes is a reasonable place to start — and set the timer. When it rings, write one line: what you actually got done, and whether the last five minutes were as good as the first five. That is the whole protocol. You are not grading yourself; you are collecting two data points per block.

After six or eight blocks you will see a pattern. If the last five minutes were consistently fine, lengthen the block by ten minutes and repeat. If they were consistently mush, shorten it. Most people land somewhere between 30 and 60 minutes for dense work and can go longer on work they already know how to do.

The reason this works better than adopting a rule is that your limit is not one number. It moves with sleep, with the difficulty of the material, and with how clearly you defined the task. What you are learning is the range and what moves it.

What the clock actually costs

This is the part most advice leaves out. A study block is not just focus time — the breaks are real minutes of your evening too, and if you plan in focus time you will plan an evening that does not fit.

The table above counts the breaks between rounds but not the one after the last round: once the final round ends you are done, and that break belongs to the rest of your day. Read it backwards when you have a fixed window — if you have two hours between dinner and bed, you are looking at four rounds, not five.

How long the break itself should be

Short blocks want short breaks. After 25 minutes, five minutes is enough to stand up, refill a glass, and look out of a window. After 50 or 90 minutes, five minutes is not enough and you will come back still tired; give it fifteen to twenty and actually leave the desk.

The one thing that reliably ruins a break is a feed. Scrolling is not rest for the same machinery you were just using — you are still reading, still deciding, still switching. Walking, eating, washing something, lying on the floor: those give the attention system something else to do. This is the part of every method that people drop first, and dropping it quietly turns a 50/10 into a 60 minute block with a worse second half.

If you are studying for an exam in a week

Short blocks win when starting is the hard part, and exam weeks are usually a starting problem rather than an endurance problem. Twenty-five minutes is a small enough promise that you will make it at nine in the morning when you do not want to.

Long blocks win when stopping is the hard part — proofs, essays, debugging, anything with a long run-up where a five minute break breaks a thread you had just picked up. If you catch yourself resenting the timer because you were finally getting somewhere, that is the signal to lengthen the block rather than to abandon timers.

More

• How long should a study session be?
• How many pomodoros is that, really?
• How many pomodoros a day?
• Ultradian rhythms: the 90 minute block, honestly
• The timer, and all 16 guides