Written by Dr. Leo Maren · Co-founder of Fwocus · Behavioural scientist · PhD, Imperial College London
A 90-minute block is for one demanding task that needs a long warm-up. Writing, programming, analysis, and hard problem sets fit. Email, admin, and small fixes do not. They fill 90 minutes without using them well.
I write in 90-minute blocks on most mornings. The first 15 or 20 minutes are slow. After that I can hold a whole argument in my head, and a chime at minute 25 would break it.
The 90-minute ultradian rhythm: what the evidence shows
Many productivity articles say that your brain runs on 90-minute cycles of focus and fatigue. The claim starts from real research. The research does not support the strong version.
The sleep researcher Nathaniel Kleitman proposed a basic rest-activity cycle, or BRAC. In his 1982 review, he argued that this cycle shows in the alternation of non-REM and REM sleep. He also argued that a similar cycle continues while we are awake.
The sleep cycle is well documented. The waking cycle is not. Neubauer and Freudenthaler (1995) tested 60 people every 10 minutes for 9 hours on a simple cognitive task.
They found no significant 90-minute rhythm in performance, alertness, mood, or heart period. Longer rhythms explained more of the variation. The authors concluded that some earlier positive reports could come from weak statistical methods.
What the evidence shows: alertness changes during the day. What it does not show: that everyone has a regular 90-minute cycle to plan work around. It also does not show that you should stop at exactly minute 90.
Before we built this page, I tested the idea on myself. For two weeks I rated my alertness every 15 minutes while I worked. I found no regular cycle. I found strong effects of sleep, lunch, and the type of task.
A better reason to work in 90-minute blocks
Better support for 90 minutes comes from how skilled people practice. In the classic violin study by Ericsson, Krampe, and Tesch-Römer (1993), diaries showed that uninterrupted practice sessions lasted between 1 and 1.5 hours.
The best violinists did about four hours of deliberate practice per day, spread over several sessions. The authors also noted that they napped more than a comparison group of music teachers. They read this as a sign that practice at that level is tiring.
This is observational evidence. It shows what dedicated students did, not that 90 minutes is the best block for you. It does suggest two useful limits: sessions of one to one and a half hours, and about four hours of hard focus per day.
In practice, plan no more than two or three 90-minute blocks in a day. On my best days I complete two. The rest of the day goes to meetings, email, and review, and I no longer count that as failure.
How to plan one 90-minute block
Choose one task with one result. Write it on the task line, for example "Draft section 3" or "Fix the import bug."
Prepare before you press Start. Open the files you need and close what you do not need. The first minutes are hard enough without a search for a document.
Expect a slow start. The first 15 minutes often feel unproductive. For me that is the warm-up, not a sign that the block failed.
Keep a note for other thoughts. Write each one down and return to the task.
Take a real break after the chime. Leave the desk. I take 15 to 20 minutes, but no study gives a correct number.
Should you pause in the middle? A stretch of a minute or two does not end a block. I stand up near the halfway point when my back asks for it, and I leave the timer running.
When 90 minutes is the wrong choice
Do you check the time left every few minutes? Then the block is too long for that task or that day. Change to the 50 minute timer or the 25 minute timer.
If you do not want a fixed end at all, count up. The deep work stopwatch shows how long you actually focused. That number is good data before you choose a block length.
When I started, my own count-up records surprised me. My longest real focus periods were closer to 70 minutes than 90. I kept the 90-minute timer anyway, because the last 20 minutes are where I review and plan the next block.
If you want 90 minutes of sleep, not work, use the nap timer. It has a 90-minute option with a wake chime that repeats. This timer plays one chime only.
How this 90 minute timer works
The timer starts at 1:30:00 and counts down in hours, minutes, and seconds. Press Start or Space to begin. A soft tone plays. Space pauses and resumes, R resets to 1:30:00, and F opens fullscreen.
The −1 min and +1 min buttons, or the Down and Up arrow keys, change the time before or during a block. The browser tab shows the time left.
At zero, a singing-bowl chime plays, the digits pulse, and "Time's up" appears. On supported browsers, the screen stays awake while the timer runs, which matters in a long block. In a background tab the time stays correct, but the browser can delay the chime.
Nothing is saved, and you do not need an account.
FAQ
Is the 90-minute ultradian rhythm real?
Partly. The cycle of non-REM and REM sleep is well documented. A regular 90-minute cycle of focus while awake has not been found reliably. Use 90 minutes because it suits long tasks, not because of a biological clock.
How many 90-minute blocks can I do in a day?
Two or three for most hard mental work. The violinists in Ericsson's study did about four hours of deliberate practice per day, spread over several sessions.
How long should the break be after 90 minutes?
No study gives a correct length. Take long enough to leave the desk and feel ready to start again. I take 15 to 20 minutes. The break timer helps you keep it to a fixed length.
Is a 90-minute block better than Pomodoro?
It is better for tasks with a long warm-up and worse for tasks you avoid. For a single short start, use the 25 minute timer.