What's up with the recommendations for little sleep?

There are several times in the suttas where sleeping too much is discouraged, and the ideal amount recommended seems quite low. I’ve heard the Buddha slept as little as 2 hours a night, more frequently I’ve heard that the ideal recommendation is 4 hours.

This seems… quite low?

I used to believe that 8 hours a night was needed. I also had trouble with insomnia at times. I then learned that, actually, 7 hours is probably the ideal amount. Since restricting myself to this amount I have had far fewer difficulties sleeping.

Here are the graphs in the paper I have linked to (unfortunately the link may be behind a paywall for some, but a google along those lines should return similar results):

Quite notably, the mortality risk for sleeping as little as 4.5-5.5 hours a night is lower than those who sleep 7.5-8.5 a night (the usual caveats about this only being an observational study apply). So, the cultural wisdom that 8 hours a night is needed seems incorrect.

Nonetheless, sleeping 3.5-4.5 hours a night is not associated with ideal health outcomes, and has been linked to poor mental health, memory, etc.

Quite plausibly, monastics who spend a large portion of their day meditating are able to get some of the effects of sleep during their meditation. It would make sense then that they need less sleep.

If this is what is going on, then the data is easy to understand. Is the recommendation for a small amount of sleep only applicable to monastics and not lay followers?

I am curious how much the monastics on this website sleep per night, although I would understand if that is too personal of a question. Finally, I know that some centers have an all night meditation once or twice every lunar cycle. This has always seemed strange to me and a type of unfruitful and painful asceticism. Was this specific practice recommended?


The recommendations are for those who are practicing ardently, with clear comprehension and mindfulness not for a general or beginning practice. With a developed meditation practice, the mind doesn’t require a lot of sleep time, I think 4 hours - or the middle watch of the night - is common at many meditation centers.

Additionally the recommendation is too avoid the sloth and laziness by means of too much sleep although a teacher once said, it is better to be asleep that to be performing unskillful deeds.


I can’t comment much on the physiology (and I’d be happy to receive more info about this), but there are a some things that come to mind.

The Buddha often talked about mental and bodily tranquility or ease that arise when cultivating deep serenity and wakefulness. The mind comes to rest from agitating thought processes, and the body becomes very relaxed, easeful, full yet light. The breath also becomes very subtle and calm. All of this seems to have an affect on the nervous system — putting the body into ‘rest & digest’ mode as opposed to an anxious ‘fight or flight’ mode that comes when the breath and body are tight, and energy lifted up near the chest and head due to mental activity and tension.

So the mind is very still and restful, and there is very low tension on the body / nervous system. I’m not sure how the quieted mental activity relates to sleep physiologically, but just how ‘sleeping on’ something allows ones mind to return refreshed and re-juvenated, samādhi has the same effect: when mental activity does return and one engages in it, the mind and body are deeply refreshed and renewed.

At the same time, and on the other end of the spectrum, mental and physical drowsiness drop away. These are of course hindrances to samādhi, and so one cultivating it will be free of them. So the mind and body will not feel the need to sleep very much, and will naturally incline to a clear, wakeful presence. This is opposed to people who deprive themselves of sleep, all the while feeling lethargic, dull, heavy, and drowsy. In that case, the body and mind clearly alert us that they lack and need sleep, as opposed to times of deep cultivation where they say the opposite.

I believe the Buddha was recommending and proposing a middle way — one that must be evaluated against what you may hear in modern Buddhist circles. If we look at a key discourse addressing drowsiness and sleep (AN 7.61), the Buddha tells Mahāmogallānā (who was struggling with sleepiness) that if he cannot resolve his sleepiness by adjusting his mind/body, then he should sleep until he is rested enough. This indicates that it is okay to sleep if the body-mind really needs it; but we don’t indulge in it or look to it immediately, because from the perspective of proper cultivation, the need for lots of sleep will fall away.

In the same sutta, the Buddha talks about the importance of maintaining restraint, inner calm, and being in secluded environments for deep practice; I believe this is related to the theme before. If the mind is active, caught up in things, and far from samādhi, it will need sleep and become drowsy; if it is wakeful yet still, the body and mind can unify and need for sleep will diminish. So he’s recommending external conditions outside of the seated session for developing wakefulness.

If we are not at this point yet, again, we don’t pretend we are and ignore our body/mind — we adjust accordingly, without going towards over indulgence. In fact, I found the article you shared interesting in saying that over indulgence in sleep has clear correlation to increased mortality. The recommended amount of, say, 6.5-7 hours is relatively moderate, and getting slightly less is still less dangerous than over-indulgence — especially if we factor in what I mentioned above about practice.

As far as the ‘watches of the night,’ it’s often interpreted to mean 4 hour periods. I’m not entirely convinced. The first watch of the night essentially means the evening. The second is the real ‘night.’ The third is the early morning before dawn — the time when monastics traditionally go alms-round (referenced in the suttas). So if the evening goes from around 6-10PM, and the dawn is around 4-5AM, we could see a rough time of about 10PM-4AM, of 6 hours of sleep, which is reasonable both in terms of bodily health and dawn-time. Even if the ‘third watch’ begins slightly earlier, the suttas say one gets up during this third watch but before alms-time, so it could be any time in the middle so long as the night is not over. They weren’t using military grade alarm clocks in Ancient India.

The point is not to go to bed early just because the day is closing, and not to stay sleeping too late so as to miss out on quiet time for cultivation before alms-round. We don’t turn the mind away from reality by taking refuge in sleep; but we don’t deny our body it’s physical necessities. The same is true for food, clothing, shelter, and medicine in the Buddha’s teaching.

I hope this is beneficial :slight_smile: That’s about most of the info on wakefulness that comes to mind, so it should be relatively comprehensive. Mettā.


One of the great monks of Thailand, now passed away, visited Bodhinyana. He said we should sleep only for 2 hours. To which a junior monk replied, “What, after lunch?”


Am curious about this since this if the first time that I have heard of this. Care to kindly share sources? I would like to read up on it :slight_smile:

I am not a monastic and understand you’re prinicipally (understandably) looking for monastic feedback here.

That said, as one who’s sat many lay retreats, I can attest to the nightime sleep needs decreasing significantly…presumably in correlation to the amount of time spent meditating during waking hours. Vaddha (above @Vaddha) put it very well, I feel.

I’ve always wondered about REM needs vs. generic sleep needs. Well, that wondering doesn’t go very far as I’m not a monastic and don’t practice seclusion as a lifestyle.

As someone with a long-time mood disorder, I’ve always ensured sufficient sleep – the 8-hour variety – to stave off the beast, so to speak. So I tend to regard admonitions around sleeping less with a grain of salt, very much in need of context, and with some light humour :grin:.


The instructions on sleep are part of the gradual training. eg:

MN39 (Sujato)
You should train yourselves like this: ‘We will be dedicated to wakefulness. When practicing walking and sitting meditation by day, we will purify our mind from obstacles. In the evening, we will continue to practice walking and sitting meditation. In the middle of the night, we will lie down in the lion’s posture—on the right side, placing one foot on top of the other—mindful and aware, and focused on the time of getting up. In the last part of the night, we will get up and continue to practice walking and sitting meditation, purifying our mind from obstacles.’

The older Horner translation of the same sutta uses the expression - ‘watch of the night’ when translating this passage, which is not, IMO, borne out by the actual pali.

MN-39 (Horner)
Thinking: ‘We must be intent on vigilance; during the day, pacing up and down, sitting down, we must cleanse the mind from obstructive mental objects; during the first watch of the night, pacing up and down, sitting down we must cleanse the mind from obstructive mental objects; during the middle watch of the night, we must lie down on our right side in the lion posture, placing one foot on the other, mindful, clearly conscious, attending to the thought of getting up again; during the last watch of the night, rising, pacing up and down, sitting down, we must cleanse the mind from obstructive mental objects’; thus, monks, must you train yourselves.

Reading the older translation and informed by western biblical sources (concepts which are unlikely to apply in the Indian context), people have made various estimations of how much time one should sleep for. This has led to the idea of 3-4 hours of sleep.

The spirit of the instructions is to practice moderation (AN6.17). However, humans have a tendency to overachieve and out do their peers. The Theravada tradition has always had an inclination to stricter ascetic practices (not finding fault - just an observation! :pray:). So we get 2 hours of sleep, all night meditations etc…

To which I might add that if following these practices, one finds their unskillful qualities decreasing while their skillful qualities increase (AN8.53) … great, Go ahead! But do keep in mind that the Buddha was not averse to resting whenever it was needed! (SN4.7)



I’m more interested in this quote: why this posture?.. Sounds really uncomfortable, and, also, what does it matter?..

Another discouese comes to mind in the Sarvāstivādin Madhyama Āgama which specifically talks about training to remain in the lions posture while asleep and review this while waking. I can’t recall the reference unfortunately. At Thag 16.10, we see a criticism of lazy slackers who sleep on their back:

With the ending of good principles and understanding, the victor’s teaching, full of all excellent qualities, has fallen apart.
This is the season for bad principles and defilements. Those who are ready for seclusion are all that’s left of the true Dhamma.

They eat until their bellies are full,
and then they lie to sleep on their backs.

The point emphasized in the early texts about sleep is to remain mindful, clear, and intending on getting up (as opposed to sinking into sleep with no thought of getting up). I believe sleeping on the side like this is meant to support this attitude of composure and clarity before and going into a sleep state, whereas rolling around and getting comfortable encourage over indulgence, day-dreaming (or ‘pre-night dreaming’ I suppose lol), etc. But this is just speculation from the material + a bit of my own experience.

Some sources talk about the benefits of being on the right side for energy channels and whatnot, but that’s not referenced in the EBTs. All in all, it encourages mindfulness of our posture/body and clarity of mind in a composed, dignified way even when going to sleep — that seems to be the main benefit as far as the Early Buddhist texts hint.

Just a public service announcement to say that reducing/eliminating sleep is an excellent trigger for bipolar disorder. And that the nature of BD is that it may very well feel like skillful qualities are increasing when it’s really just the onset of a manic episode.


The origin story to Pacittiya 5 gives a counter example:

At one time the Buddha was staying at Āḷavī at the Aggāḷava Shrine. At that time the lay followers were coming to the monastery to listen to the Teaching. When the instruction was over, the senior monks went to their own dwellings, but the newly ordained monks lay down right there in the assembly hall together with the lay followers—absentminded, heedless, naked, muttering, and snoring. The lay followers complained and criticized them, “How can the venerables lie down absentminded, heedless, naked, muttering, and snoring?”

So, it’s the rule for the mendicants, not the laity?.. And it’s just because of propriety, not because of some energies or other magic rituals?..

Bikkhu Bodhi relates the Buddha’s daily schedule:



This paper is relevant: https://uknowledge.uky.edu/cgi/viewcontent.cgi?article=1037&context=biology_facpub

Very small sample size (7 individuals) but it compared people with more than three years of experience in meditation who meditated a few hours a day to a control group. The average time of meditation for the meditators was 2.3 hours a day and their average sleeping time was 5.2 hours, as compared to 7ish hours for the non-meditators. The meditators showed no sign of cognitive deficit or adverse health indicators of sleep deprivation.


Why do you think that? As far as I know, Horner’s “watch” is a quite uncontroversial translation of yāma. Bhikkhu Bodhi uses the same in his SN and Sn translations.

Pali English Dictionary - yāma

And an old thread on the subject:

Three Watches


Just my personal opinion! I do not think that timekeeping was sufficiently advanced in the Buddha’s time to be able to measure time specifically enough to be able to speak of ‘a watch’. For example, there are many suttas which have the Buddha robing up for alms round and arriving at a village or town only to realize that ‘its too early to wander for alms in …’ and then proceeding to spend a few hours in discussion elsewhere - which would have been unlikely if time was being measured with reasonable accuracy (as implied by water clocks, sun dials etc). Timekeeping specific enough to be able to divide the night into hours (as is implied by the expression of ‘a watch of the night’) seems IMO to be a later invention.

Hence I feel that a fairer translation might be ‘part of the night’.

Still, just my personal opinion!


Well, the idea of keeping time is an early one, but, yeah, I don’t think that Buddha and his Sangha had any access to sun dials, water clocks or fire clocks. People can guesstimate time by looking at the sun and shadows - and they can be pretty accurate in their guesses! But it’s more difficult at night.
Western monks ofter kept time at night by reading specific texts (The Book Of Hours) and ringing a bell. I wonder if there were anything like this in Early Buddhism?..

I don’t think the word ‘watch’, used here for the Pali word ‘yāma’ refers to a clock or ‘wrist watch’ but rather a general observation about the progress of time through the night.

The term is used in bible translations and likely adopted from there.


This is exactly my point! The term refers to an ability to divide the night reasonably accurately and reflects technological skills from a period and culture around 500 or even more years after the times of the EBT.

It is precisely this implication of accuracy that leads us to the conclusion that the middle watch of the night is a period of 4 hours. Which then fosters a practice where the person sleeps far too little than is good for their health.

However, translating the term as ‘part of the night’ removes these hidden assumptions of (i) reasonable accuracy of measurement of the middle part of the night as being 4 hours and (ii) the night being neatly subdivided into 3 equal parts.

This leads us to be more mindful of the physiologically sufficient period of sleep required by most people (around 6-7 hours ?) while still maintaining the spirit of the injunction to sleep less and practice more.

Still, as I said earlier - this is just my opinion! (And, no I’m not attached to it. :rofl: )


It’s not clear to me why those described in the Old Testament would be better able to divide their nights into roughly 3 parts than people in India.
Or why ‘part’ is better than ‘watch’ to describe a portion.
Seems a question of taste.