A polyphasic sleep schedule splits sleep into several blocks across 24 hours, from Everyman’s four hours down to Uberman’s two. The split patterns with real evidence behind them keep total sleep close to a normal night. When a 2026 lab study tested Uberman directly, volunteers slept worse and stayed less alert than on one short block.

Van Gogh's The Siesta: two farm workers asleep in the hay at midday, the split day of sleep behind the oldest polyphasic sleep schedule
Vincent van Gogh, The Siesta (after Millet), c. 1890. Musée d’Orsay, Paris. Public domain, via Wikimedia Commons.

Why the usual advice fails

Search for polyphasic sleep and you get the same page five times. A list of schedules, a box of claimed benefits, a box of risks, and a doctor saying don’t. All of it is correct. None of it is much use.

The problem is that “polyphasic” covers two very different things. One is splitting a normal amount of sleep into pieces: a night with a gap in it, or a night plus a nap. The other is cutting sleep to two or four hours and spreading the crumbs around the clock. The first has history and some evidence behind it. The second has enthusiasts, apps and, until recently, almost no lab data.

Lump them together and you get vague warnings that people ignore. Split them apart and the answer gets sharp. So here are the schedules, oldest first, with where each one actually comes from and what the people who measured it found.

What the sleep researchers actually found

Schedule Pattern Sleep per 24 hours Where it comes from What the evidence says
Segmented (“first and second sleep”) One night, split by a waking gap A full night Pre-industrial Europe, per historian Roger Ekirch Disputed by historians; appears in a lab under very long nights
Biphasic (night plus nap, siesta) Shorter night, one daytime nap Close to normal Long tradition; tested in teenagers in 2019 Mixed: sharper mind, worse blood sugar
Racing sailors Naps of 20 minutes to 2 hours About 6.3 hours Claudio Stampi’s 1989 study of 99 sailors Short sleep went with faster races, in an emergency setting
Dymaxion 30 minutes about every 6 hours About 2 hours Buckminster Fuller, reported in Time, 1943 One man’s self-report
Uberman Six 20-minute naps, one every 4 hours 2 hours Named online around 2000, per a fan wiki Tested in 2024 and 2026: people quit or did worse
Everyman About 3-hour core plus three 20-minute naps About 4 hours Polyphasic hobbyists; origin undocumented No controlled trial found
Triphasic Three short blocks, as Cleveland Clinic describes it 4 to 5 hours Polyphasic hobbyists; origin undocumented No controlled trial found

Now the evidence, one row at a time.

Segmented sleep: real history, open argument. The historian A. Roger Ekirch argued in the American Historical Review in 2001 that pre-industrial Westerners slept in two shifts. In a 2024 reply in Medical History, he sums it up: most people went to bed between 9 and 10pm, woke from their “first sleep” about midnight, stayed awake for up to an hour or so, then took a “second sleep” until dawn. He says he has found upwards of two thousand references to it, and that the pattern faded with the Industrial Revolution. Not everyone buys it. Niall Boyce of Birkbeck, University of London, re-read the sources in Medical History in 2023 and concluded that segmented sleep “may have been practised by some individuals” but was not the sole or even necessarily the main pattern. And when Yetish and colleagues measured sleep in three pre-industrial societies for Current Biology in 2015, they found nights “not interrupted by extended periods of waking”. So treat segmented sleep as disputed.

The lab has one striking result. In 1992 Thomas Wehr, at the US National Institute of Mental Health, moved volunteers from 16 hours of light a day to 10, meaning 14 hours of darkness. Their sleep “usually divided into two symmetrical bouts”, with a one-to-three-hour waking gap between them. Ekirch notes that it took several weeks without light at night for the pattern to appear. The point for you: a split night is a normal-length night with a gap in it. Nobody involved slept less.

The nap: helps one thing, hurts another. In 2019 June Lo and colleagues in Singapore gave sleep-restricted teenagers, aged 15 to 19, the same 6.5 hours a day in two ways. One group got 6.5 hours at night. The other got 5 hours at night plus a 90-minute afternoon nap. The split group had fewer lapses of attention, better working memory and better mood. They also had bigger rises in blood sugar on a glucose test. Both groups did worse on attention than teenagers given nine hours. The authors’ advice: get the recommended amount of night-time sleep.

The sailors: the original science, often misquoted. The researcher Claudio Stampi published a field study of 99 sailors in solo and two-person ocean races in Work & Stress in 1989. As a 2026 paper summarises it, most slept in episodes of 20 minutes to two hours, about 6.3 hours a day in total, and shorter sleep went with better race results. Note the number. That is six hours, not two, and it was a race where sleeping long could sink the boat.

Dymaxion: one inventor’s word. Time reported on 11 October 1943 that Buckminster Fuller napped for half an hour at the first sign of fatigue, which came about every six hours. He claimed two years on it and quit because his business associates “insisted on sleeping like other men.” Nobody measured anything. We cover his story and the other people who tried this in what happened to the people who tried the Uberman schedule.

Uberman: finally tested, and it lost. Two recent studies in the journal Sleep did what blogs never could. In the first, Yevgenia Rosenblum, Martin Dresler and colleagues put ten young, highly motivated volunteers on a 20-minute nap every four hours for a planned eight weeks. All but one quit within the first month. In a 2026 editorial the authors say the volunteers left because of the “strong impact on their social life and subjective wellbeing”. The one who lasted, a 25-year-old man, managed five weeks, and his growth hormone release fell by more than 95%.

In the second, published in 2026, Tiffany Koa and June Lo at the National University of Singapore gave 40 healthy young adults either Uberman or a single two-hour sleep, with brain monitoring during every sleep slot. Same total sleep opportunity, different shape. Both groups were sleepier and less alert than well-rested people, which is no surprise. The surprise was that Uberman was worse than one short block. The polyphasic group got less actual sleep, took longer to drop off, woke more and had poorer vigilance, especially in the morning. Across 24 hours they averaged 18.2 minutes of deep sleep and 7.4 minutes of REM. Advocates say the naps squeeze in all the deep and REM sleep you need. The authors say those amounts are too short to back that up.

Honest limits: the 2026 study ran for one 24-hour period, and the authors say so. Fans will argue that adaptation takes weeks. But the 2024 study tried weeks, and nine of ten motivated volunteers could not get there.

Everyman and triphasic: untested. I could not find a controlled trial of either. The 2021 National Sleep Foundation panel led by Matthew Weaver screened 40,672 publications on polyphasic schedules, kept 22, found no evidence of benefit, and linked these schedules to worse physical health, mental health and performance. Both still leave you three hours or more below the seven to nine that the US National Heart, Lung, and Blood Institute recommends for adults.

Why splitting sleep works for some and fails for most

This is my reading of the studies above.

Total hours decide it, not the number of pieces. Every version with some support keeps sleep near a normal amount: Ekirch’s two sleeps, Wehr’s long winter nights, the teenagers’ nap. The versions that fail are the ones that cut the total to two or four hours. Splitting is not magic and not poison. Cutting is the problem.

Short naps are mostly light sleep. The 2026 lab found that 20-minute slots fill up with falling asleep, waking and light stages. Deep sleep and REM barely got a look in. The nap is a good top-up for a full night, and a poor replacement for one.

The rest of the world runs on one block. Fuller’s associates slept like other men. The 2024 volunteers quit over their social lives. A schedule that fights everyone else’s clock has to win every single day, and most people’s lives will not allow that.

The sailors are a special case. Stampi’s sailors split their sleep because the boat demanded it, for the length of a race. That is a survival tactic for a few weeks, not a lifestyle.

How to try a polyphasic sleep schedule safely this week

This is Copy John’s adaptation, not a plan from any of the studies. It lets you test the one kind of split sleep that has evidence on its side, without cutting your total.

  1. Write down what you want. More waking hours, or more alertness in the afternoon? If it is hours, find them elsewhere first. Our Uberman piece shows how one tester found he slept more in total, not less.
  2. Fix your total at seven to nine hours. That is the NHLBI range for adults. Any schedule you try this week has to add up to at least seven.
  3. Fix your wake time. Same time every day, with weekends within about an hour, per the NHLBI’s healthy sleep habits.
  4. Add one nap, not six. No more than 20 minutes, earlier in the afternoon, again per the NHLBI. That is a biphasic day, the mildest form of polyphasic sleep.
  5. If you wake in the night, don’t panic. If you wake for a while around midnight and drift off again, Ekirch would say people have done that for centuries. If it happens most nights and you are tired by day, talk to a doctor.
  6. Score yourself twice a day. At 10am and 3pm, rate your alertness from one to five. Note your total sleep. Nothing else.
  7. On Sunday, do the sums. Uberman promises that going from 7.5 hours to 2 frees 5.5 hours a day, or 38.5 a week. Compare that with your scores, and with the 2026 finding that those hours come with worse attention. Then decide what an extra hour is worth if you spend it half asleep.

If you feel exhausted on a normal night’s sleep, that is a question for a doctor, not for a new schedule.

What not to copy

The adaptation all-nighter. Some adaptation plans, like the one Charles Lee followed in 2016, start with a day or two awake to force your body onto the naps. That is deliberate sleep deprivation, and it is the worst state in which to drive or look after anyone.

The “you get all your REM” claim. The 2026 lab measured it. 7.4 minutes of REM in a day is not all of anything.

The sailors’ numbers. People cite Stampi to justify two-hour days. His sailors averaged about six hours, under race conditions.

The calculator. A schedule maker can space your naps neatly. It cannot change how much sleep you actually get, which is the only number the studies care about.

The famous name. Leonardo da Vinci’s name is stuck to Uberman with no source at all. We traced that in the Da Vinci sleep schedule is made up.

Frequently asked questions

Is polyphasic sleep better than monophasic?

Not on current evidence. In the 2026 Sleep study by Koa and Lo, Uberman produced less sleep and worse morning vigilance than one block of the same length. The 2021 National Sleep Foundation panel found no evidence of benefit from polyphasic schedules.

Does polyphasic sleep work?

As a way to sleep much less, no study shows it working. Nine of ten committed volunteers quit an Uberman trial within a month in 2024. Splitting a normal amount of sleep, such as a night plus a short nap, is a different thing, with mixed but far milder results.

Is a polyphasic sleep schedule good for students?

The best test is Lo’s 2019 study of 15 to 19 year olds. Splitting 6.5 hours into a night plus a nap helped attention and mood but raised blood sugar more. The authors’ advice was to get the recommended night-time sleep instead.

What polyphasic sleep schedule does Ronaldo use?

Nobody has shown one. The “five 90-minute naps” story has no first-hand source; I found it only on fan and aggregator sites. Nick Littlehales, the sleep coach widely reported to have worked with him, counts sleep in 90-minute cycles, about 35 a week, in his 2016 book Sleep. Five cycles a night is 7.5 hours. My guess is that is where “five naps” came from.

Sources


This is part of Copy John’s look at famous sleep schedules, which starts with the Da Vinci sleep schedule. More in the routines collection.

Get one of these a day, free →

More like this