Extended Fasting: 24, 36, 48 and 72 Hours Compared

Quick answer: Extended fasting means going longer than a day without food. The useful range for someone doing this without medical supervision is 24 to 72 hours, and longer is not better — the benefits flatten out while the costs keep climbing. For most people the honest recommendation is a weekly 24-hour fast, an occasional 36 or 48-hour one, and 72 hours rarely if ever. This page compares the tiers, sets out what each one costs, describes what a long fast actually feels like, and answers the refeeding question in full.

Every extended-fasting question eventually collapses into one: how long?

The answer that gets clicks is "longer". The answer supported by what has actually been measured in humans is more boring and more useful, and it is below.

What counts as extended fasting

Anything past roughly 24 hours. Below that you are doing time-restricted eating — 16:8, 18:6, 20:4 — where you eat every day and vary the window.

The line matters because the two are different activities. Daily fasting is a habit you build. Extended fasting is an event you plan: it needs preparation, electrolytes, a clear day, and a way to end it. It also needs a screen first, which is the next section.

The screen that comes before any of this

Extended fasting is optional for everyone. If any of the following is true of you, the right next step is a conversation with your doctor, not a shorter fast:

Diabetes, or any medication that lowers blood sugar
Any prescription medication that must be taken with food
Pregnancy or breastfeeding
Any history of an eating disorder, current or past
Under 18
Being underweight, or losing weight without meaning to
A heart, kidney or liver condition, or a history of fainting or low blood pressure

The extended fast planner runs the same screen before it draws anything, and a hit stops it rather than warning you. Nothing on this page is worth a hypoglycaemic episode or an eating-disorder relapse.

The tier map

Length	Who it suits	What it buys	What it costs	Sensible frequency
24 hours	Anyone comfortable with daily fasting	Full switch to fat and ketone fuel; a real weekly deficit	One awkward day; hunger waves	Weekly
36 hours	People who found 24 easy	Longer time in deep fasting metabolism; two nights of sleep inside the fast	A full waking day plus a morning; more electrolyte demand	Weekly to monthly
48 hours	Experienced fasters, occasionally	Two full days of deep fat and ketone metabolism	Light-headedness, poor sleep, no hard training for two days	Monthly at most
72 hours	Few people, rarely	Deeper metabolic stimulus; the experience itself	Measurable muscle protein breakdown; three days off training	Once or twice a year, if at all
Beyond 72	Nobody, unsupervised	Nothing this site can stand behind	Unknown, unmonitored	Not without a clinician

24 hours

The workhorse. One meal to the same meal the next day, once a week, which is exactly the structure of Eat Stop Eat — set out in full further down this page. Long enough that the metabolic switch Anton and colleagues describe is well behind you, short enough to slot into an ordinary week without rearranging anything.

The direct evidence at this length is decent for a fast this short: a randomised cross-over trial in 30 healthy volunteers found a single day of water-only fasting acutely raised growth hormone and HDL cholesterol and lowered triglycerides — with most markers back at baseline within 48 hours. That last part is the point. A 24-hour fast is a stimulus you repeat, not a switch you flip.

Start here. Every other tier on this page assumes you have already done this one.

36 hours

Dinner one evening to breakfast two mornings later. It buries two nights of sleep inside the fast, which is why many people find 36 hours barely harder than 24. It is the natural next step, and the rhythm inside alternate day fasting.

48 hours

Two full days. This is where the fast stops fitting around your week and your week starts fitting around the fast.

It is also a duration researchers have used, so there is direct human data — including magnetic resonance measurements of triglyceride moving into the livers of men and the muscles of women across a 48-hour fast, while the whole-body glucose and oxidative responses stayed identical between the sexes.

Worth doing occasionally if you are experienced. Not worth doing weekly.

72 hours

Three days, and the first tier where a real physiological cost is measurable rather than theoretical. In eight healthy men fasted for 72 hours, net phenylalanine release from skeletal muscle rose and mTOR signalling fell by roughly half — muscle breakdown outpacing synthesis.

Meanwhile the benefits most people are chasing at this length — the autophagy and immune claims — have not been quantified in humans at any specific hour. That combination, a measured cost against an unquantified benefit, is why this tier is rare rather than aspirational.

What three days actually feel like is set out in the day-by-day walkthrough below.

Beyond 72 hours

This site stops at three days, and the planner will not draw anything longer.

The reason is in the safety data. The reassuring numbers on multi-day fasting — under 1% adverse effects across 1,422 people fasting 4 to 21 days, and 75% mild adverse events across 768 water-only fasting visits — come from supervised programmes. The 1,422-person cohort was not water fasting at all: participants took 200 to 250 kcal a day and saw a physician two or three times a week. That evidence says long fasts are well tolerated when someone is checking the bloods. It says nothing about doing the same thing alone at home.

So how long should I fast?

Here is the honest version, and it will disappoint anyone hoping the answer is a bigger number.

Length buys you depth, not multiples. A 72-hour fast is not three times as useful as a 24-hour one. The big metabolic transition — glycogen down, fat and ketones up — is largely complete inside the first day. Everything after that is time spent in a state you already reached, and the extra hours add cost faster than they add benefit.

Frequency beats duration. A weekly 24-hour fast gives you 52 fasts a year. A quarterly 72-hour fast gives you four. Which of those do you think moves your body composition and your relationship with food more?

Nothing here beats ordinary dieting on the scale. The most rigorous long-run test of a fasting protocol — a one-year randomised trial of alternate-day fasting against daily calorie restriction — found no advantage for fasting on weight loss at 6 or 12 months, and the fasting group had the highest dropout rate at 38%. Fasting works roughly as well as careful daily restriction. Its advantage is that some people find it far easier to keep doing, which is a real advantage, and not the same as a metabolic one.

Your metabolism is not the reason to worry. It does not fall during a short fast. Measured across 84 hours of starvation in lean healthy people, resting energy expenditure rose rather than fell, tracking a large increase in noradrenaline. The metabolic slowdown people fear belongs to months of undereating, not to two days without food.

So: weekly 24 hours, occasional 36 or 48, 72 rarely. If you want a different answer, the honest one is that the evidence does not support paying more for it.

Eat Stop Eat: the weekly 24-hour fast under a name

That weekly 24-hour fast already has a name. Brad Pilon set it out in 2007, out of graduate nutrition research at the University of Guelph, against an industry then insisting on six small meals a day. Eat Stop Eat is one or two 24-hour fasts a week and normal eating the rest of the time. No windows, no counting, no rules on the other five or six days.

The detail people get wrong is that the fast runs meal to meal, not across a calendar day. Dinner Monday to dinner Tuesday means you eat on both days — you skip the meals in between. That matters more than it sounds. "I am not eating today" is a much harder sentence to live inside than "I am eating tonight, I just have to get to dinner."

Dinner-to-dinner is the version most people find easiest, because only one waking day sits inside the fast and you never wake up already hungry. Lunch-to-lunch behaves the same way. Breakfast-to-breakfast is the hardest: it asks you to go to bed with the fast still running and then face a whole morning on top of it.

Four rules make it work:

Never two days running. Move the day around your week instead — pick one you are too busy to eat on and the fast half runs itself.
Two a week is the ceiling. Past that you are on alternate day fasting, which is a different commitment with its own dropout problem.
Not on hard training days, and not on days with a meal that matters to you.
Eat properly on the other days. Chronic undereating with a weekly fast bolted on is the one way to make this go wrong.

Break it with a normal dinner rather than a compensatory feast. You will make the calories back over the following days without forcing it.

What every extended fast needs, whatever the length

Electrolytes, on a schedule. Falling insulin makes the kidneys flush sodium, with potassium and magnesium going with it. That single mechanism explains most of what goes wrong on a long fast — the headaches, the cramps, the fog, the head-swim on standing. Get amounts from the fasting electrolyte calculator and read electrolytes during fasting for the why.

A plan for ending it. The gut quietens down during a long fast, and a huge first meal is the fastest route to feeling awful. Small, easy, protein-and-fat first; wait; then a normal meal. The how to break a fast tool builds the sequence around your actual fast length.

An honest read on refeeding risk. For a well-nourished adult, fasts in the 24 to 72-hour range do not reach the criteria clinicians use to flag high risk. If you are underweight, have been eating badly for a while, drink heavily, or have any history of an eating disorder, the risk is real. The section below gives the full criteria and what to do about them.

A stopping rule. Faint, confused, breathless, or a racing or irregular heartbeat means eat now, regardless of the timer.

What a long fast feels like, day by day

The hour markers describe your metabolism. They do not describe your day, and the two diverge more the longer you go.

Day one is a 24-hour fast and behaves like one. Hunger arrives in waves at your usual mealtimes and passes in twenty minutes or so if you sit through it. It does not build steadily; that is the single most useful thing to know before your first long fast.

Night one is where some people first notice a cost. Sleep can go light or broken. Rising noradrenaline is the likely reason, and it is the same signal holding your metabolic rate up rather than letting it fall. Magnesium in the evening, a cool room and no caffeine after mid-morning are the practical fixes.

Day two is usually the hardest, and hunger is rarely what makes it hard. It is the head-swim when you stand up, cold hands, a flat mood and a shorter fuse. Stand up slowly. Keep the salt going. Cancel anything needing sharp physical judgement, and do not drive if you feel at all faint.

Day three, on a 72-hour fast, is often reported as easier than day two — steadier, sometimes unusually clear-headed. Do not read that as permission to keep going. Feeling good at hour 60 tells you nothing about what your blood chemistry is doing.

Through all of it, people report unusual mental clarity and people report the exact opposite, often on the same day of the same protocol. Neither is a signal about whether the fast is working.

Refeeding syndrome: who is at risk, and what to do

Refeeding syndrome is a collapse in blood chemistry that happens when someone who has been starved for a long time starts eating again. The danger is in the food, not in the fast. It is real and it kills people, and it belongs to a specific group — so the answer has two halves, and both matter.

What goes wrong. Through a long starvation the body runs on fat and ketones and insulin sits at a floor. The mineral stores inside cells run down while the blood test still reads normal, because the body pulls minerals out of cells to hold blood levels steady. The tank is empty; the gauge reads full. Then food arrives. Carbohydrate drives a surge of insulin, insulin moves things into cells, and phosphate, potassium and magnesium go in with it — falling fast, from a normal-looking baseline, within hours of the first real meal. Phosphate is the one that kills. It is the P in ATP, so the tissues that spend the most energy fail first: heart muscle and the diaphragm. Thiamine is the other half. Insulin does not drive it anywhere; the body simply burns through it metabolising the carbohydrate, and a starved person has usually run that store down already.

Who is at high risk. NICE publishes the checklist clinicians work from. Any one of these is enough: a BMI under 16, unintentional weight loss of more than 15% in the last three to six months, little or no intake for more than 10 days, or low potassium, phosphate or magnesium before feeding starts. So is any two of these: a BMI under 18.5, unintentional weight loss of more than 10% over three to six months, little or no intake for more than five days, or a history of alcohol misuse or of drugs including insulin, chemotherapy, antacids or diuretics.

Read the list again and notice what dominates it. Not the length of one fast, but the state the person was already in.

Where an extended fast sits. A well-nourished adult at a normal weight, eating a full diet up to the moment the fast starts, scores nothing on the first list and nothing on the second. Seventy-two hours is not five days, and three days is nowhere near long enough to empty the intracellular phosphate and thiamine stores the syndrome depends on. The panic attached to a weekend fast is a hospital problem moved onto a population it was never described in.

What to do if the criteria describe you. Talk to a clinician before you fast, not before you eat again — and treat that as the answer rather than a gentler first meal. Where the risk is real it is managed medically: bloods checked, feeding restarted at a maximum of 10 kcal per kilogram a day — as little as 5 in extreme cases — and built up to full needs over four to seven days, thiamine given before and through the first ten days, and potassium, phosphate and magnesium supplemented from the start rather than corrected first. Ten kcal per kilogram is a ceiling, not a floor, and it is roughly a third of what the same guideline gives somebody who is not at risk. That is a treatment plan, not a diet tip.

What to do if they do not. Break the fast small anyway. A few hundred calories of protein and fat, an hour's wait, then a normal meal. A large first meal after three days lands badly on a gut that has gone quiet. The bloating and the energy crash that follow are not refeeding syndrome, and they are still worth avoiding — the how to break a fast tool builds the sequence around the fast you actually did.

Stop and get medical help if you become confused, breathless or faint, if your heart races or beats irregularly, or if your legs swell after you start eating again. None of those is a symptom to push through.

Where the hour markers come from

Every fasting chart, including the zone model on this site, puts hours on the stages of a fast. Some of those are well measured in humans — the metabolic switch past roughly 12 hours, the rise in ketones, the rise rather than fall in metabolic rate. Others, autophagy above all, are models built from animal work and human signalling proxies rather than direct human measurement.

That distinction is worth carrying into any fasting content you read, including this site's. A number on a chart is not the same kind of claim as a number from a trial.

Choosing a rhythm you will keep

If you are new to this, a sensible year looks like: build a comfortable daily fasting habit first, add a weekly 24-hour fast once that is boring, try a 36-hour fast after a couple of months, and only then consider 48. Most people never need to go further, and what actually changes week by week is a better guide to whether it is working than any single long fast.

The extended fast planner will lay out any of these hour by hour, with electrolyte checkpoints, sleep and training placement, and a refeeding section it will not let you skip.

How Fasted helps

Fasted runs the timer for any of these lengths with the metabolic phase shown as you cross it, so a long fast has landmarks rather than just a rising number. Electrolyte logging and reminders keep the minerals on schedule, weight logging gives you the before and after, and every extended fast lands in the same history as your daily ones — which is the only way to see whether the long ones are earning their place.

Frequently Asked Questions

What is the best length for an extended fast?

For most people, 24 hours done weekly. It reaches full fat and ketone metabolism, costs one awkward day, and can be repeated fifty times a year — which matters far more than depth. Move up to 36 or 48 hours occasionally if you want a bigger stimulus, and treat 72 hours as a rare event rather than a goal.

Is a longer fast always better?

No. The major metabolic transition is largely complete within the first day, so extra hours extend a state you have already reached rather than unlocking a new one. Meanwhile the costs keep rising: by 72 hours, muscle protein breakdown is measurable in humans. Longer fasts buy depth at an increasing price, and past three days they belong under medical supervision.

How often can I do extended fasts?

A 24-hour fast weekly is sustainable for most healthy adults. A 36-hour fast fits weekly to monthly. Forty-eight hours is a monthly ceiling at most, and 72 hours is best kept to once or twice a year. Doing long fasts frequently makes it hard to eat enough protein and repeats the muscle cost without evidence of extra benefit.

Do I need electrolytes on every extended fast?

On anything past about 24 hours, almost certainly. When insulin falls the kidneys excrete more sodium, and water, potassium and magnesium follow, which is why headaches, cramps, dizziness and fatigue are usually mineral loss rather than hunger. Zero-calorie electrolytes do not break a fast, and taking too little is the most common reason a long fast goes badly.

Does a 72-hour fast reset your immune system?

No, and the study behind that headline does not claim it does. Its regenerative findings — stem cells shifting toward renewal after prolonged fasting — were in mice. The human part was preliminary data on lymphocytes in cancer patients fasting around chemotherapy, under clinical care. That is interesting science. It is not evidence that a healthy person fasting three days at home rebuilds an immune system.

Can you get refeeding syndrome from a 48 or 72-hour fast?

For a healthy, well-nourished adult, no. Two or three days meets none of the published criteria that mark someone as high risk, and it is not long enough to drain the mineral stores the syndrome depends on. The answer flips entirely if you are underweight, have been eating little for a while, drink heavily, or have any history of an eating disorder — in that case speak to a doctor before you fast at all, not just before you eat again.

Is extended fasting safe without a doctor?

For a healthy adult with none of the contraindications, fasts up to 72 hours are widely done alone and the published safety data is reassuring — but that data comes from supervised programmes, so it supports a careful solo attempt rather than proving one is safe. If you have diabetes, take medication with food, are pregnant or breastfeeding, are under 18, or have any history of an eating disorder, extended fasting is a medical decision and not a lifestyle one.

Does extended fasting slow your metabolism?

Not over the durations covered here. Measured through 84 hours of starvation in lean healthy people, resting energy expenditure rose rather than fell, alongside a large increase in noradrenaline. The metabolic adaptation people worry about comes from sustained undereating over months, and the way to avoid it is eating properly between fasts rather than fasting for less time.