Ultradian rhythms: the 90 minute block, honestly

What it costs on the clock

One full cycle is 7h 20m. The break after the last round isn’t counted — you’re done.

Where the 90 minutes comes from

Nathaniel Kleitman, who co-discovered REM sleep, described a cycle of roughly 90 minutes running through sleep. He proposed that a related cycle continues during waking hours — the basic rest-activity cycle. That proposal is the origin of every '90 minute focus block' article you have read.

The sleep cycle is well established. The waking version is much less so: estimates of its length vary widely between people and studies, and whether it maps cleanly onto the ability to concentrate on a specific task is not settled. Treating it as a fixed personal timer is going considerably past the evidence.

What is probably true anyway

Two things survive the caveats. Attention does fluctuate through the day in waves rather than draining steadily, and most people do get a noticeable dip in the early afternoon. Working with those waves rather than against them is sensible regardless of the exact period.

And a long block does suit some work. Anything with a long warm-up — a proof, a draft, a debugging session — pays a real cost every time you stop, and a 25 minute round can mean you never get past the warm-up. If that is your experience, a 90 minute block is worth trying for that reason, not because of the biology.

What 90/20 costs on the clock

The trade is in the break. A 90 minute block wants a real break — twenty minutes, away from the desk — and that is a third of an hour per cycle. The table above shows what that does to your afternoon for the usual targets.

Per hour of focus, 25/5 and 50/10 each spend about twelve minutes on short breaks; 90/20 spends about thirteen, which is comparable. The difference is not the total break time, it is the shape: fewer, longer interruptions, which is better for deep work and worse for fitting around a timetable.

When the long block goes wrong

Two failure modes are common. The first is that a missed start costs ninety minutes instead of twenty-five — if starting is your hard part, long blocks punish you much harder. The second is that the last twenty minutes become theatre: you are at the desk because the timer has not rung, not because you are working.

If you notice either, the fix is not to push through. Shorten the block to the length you actually sustain and lengthen it again when that stops being a struggle. The number is supposed to serve the work.

Using the dip instead of fighting it

Whatever the exact period, most people can name the hour of the day when nothing goes in. Scheduling your hardest block against that hour is a reliable way to make a method look like it failed.

Put the mechanical work there — transcribing notes, making cards, organising files — and move the hard block to the hour where you know you are sharp. This single change does more for most people than any choice of interval.

What the day usually looks like

Most people have a two to three hour window in the morning where hard material goes in, a dip in the early afternoon, and a partial recovery in the evening. The exact hours vary a lot, and a minority are genuinely reversed.

You can find yours in a week without any equipment. Three times a day, write one word for how sharp you feel and what you were doing. After a week the pattern is usually obvious, and it is more useful than any general claim about 90 minutes.

Why the exact number matters less than the shape

Whether your cycle is 80 minutes or 110, the practical advice is identical: work until the quality drops, then stop properly rather than pushing. The number is a planning convenience, not a measurement.

This is worth saying because a lot of effort goes into finding the perfect interval, and the people who get the most out of timed work are usually the ones who settled on something approximately right and then protected their breaks.

Combining long and short blocks in one day

A common pattern that works: one 90 minute block in the sharp window for the hardest thing, then 25 minute rounds for everything else. The long block gets the depth; the short rounds get the rest done without demanding much.

The mistake is the reverse — spending the sharp window on short rounds of admin because it is easier to start, then trying to do the hard thing in a long block at four in the afternoon. That is how a good method produces a bad day.

Testing it against yourself in a week

Run three days at your usual interval and three days at 90/20, on comparable material, and record one thing per block: whether the last ten minutes were as useful as the first ten. That is enough to tell you which shape suits the work you actually do.

What you are looking for is not which felt better — long blocks usually feel more virtuous — but where the quality held. Those are different questions and only the second one matters.

More

• What to actually do during the break
• How long should you study before a break?
• How many pomodoros is that, really?
• How long should a study session be?
• The timer, and all 16 guides