Every gym has the guy who brags about lifting on three hours of sleep. Seventy-seven studies just put a number on what that bad night actually costs.
A bad night barely touches how much you can lift. It wrecks how accurately you can move. Same kind of night, two completely different prices.
Every gym has one. The bleary-eyed regular who rolls in, announces a three-hour night, and loads the bar anyway like it's a badge of honor.
The unspoken rule of lifting culture is that a bad night is a test of will. Push through and you pass. Skip and you're soft. Discipline is supposed to beat fatigue every time.
In 2022, a team of sports scientists at Griffith University in Australia decided to stop guessing and put a number on it. They pulled together seventy-seven controlled studies — 959 people in all. Each one was measured doing real physical tasks on a normal night's sleep, then again on a short one. The goal was simple: work out exactly what a rough night costs.
Almost nine in ten of those people were men, all of them adults, and every one was tested after a single short night, not a long stretch of them. That matters, and it comes back up later.
The answer came back at about 7.5% worse on average. Every type of exercise they tested got worse on short sleep. But that tidy average hides the real story, because the damage was wildly uneven. Some things barely flinched. One thing fell off a cliff.
A bad night doesn't just make you feel weak — it measurably lowers performance on average, and the price isn't the same for every exercise.
- Across 77 studies, short sleep cut exercise performance by about 7.5% on average — and every type of exercise tested got worse.
- The hit lands unevenly: skill and accuracy drop the hardest, while maximal strength is the most protected.
- After a night without sleep or an early alarm, performance slips by roughly 0.4% for every hour you've been awake, so the later you train, the bigger the cost.
- Waking up early hurt performance clearly; going to bed late showed no clear loss overall, though skill tasks still fell.
- Morning workouts held up better than afternoon and evening ones after a short night.
Your Deadlift Is Fine. Your Jump Shot Is Not.
Start with the two extremes, because they tell you almost everything. Maximal strength — your heavy single, your one-rep max — dropped less than 3%. Skill-based tasks, the kind that live or die on precision and coordination, dropped almost 21%.
That's roughly a sevenfold gap. The same bad night that barely touches how much you can lift wrecks how accurately you can move.
Picture anything that runs on precision — free throws, a serve, a target. If you normally nail eight clean attempts out of ten, a drop that size takes you down closer to six. Your legs can still move the bar. It's the aim and fine control that fall apart.
The other categories landed somewhere in between — power, sprint work, intervals, endurance, muscular endurance. Every one of them got worse. But the headline is the spread. Single maximal efforts held up best. The skill tasks, which ask the most of the brain, fell furthest.
That's the part that should reframe the whole "just push through it" argument. Your training partner who swears a bad night never dents the numbers might be nearly right — about a single max effort. That was never the thing that actually broke. And after a night without sleep or an early alarm, the bill grows the longer the day runs.
The Clock Is Charging You by the Hour
Here's where the study turns into something you can actually use. The researchers lined performance up against how long people had been awake. After a night without sleep or an early alarm, performance slipped by about 0.4% for every extra hour spent awake. After a late bedtime, it did not.
Run that on a real schedule. The paper's own example: someone up at 3 a.m. who performs twelve hours later, around 3 p.m., can expect roughly a 5% loss.
Now run it on a day that started with an early alarm. Up at 5 a.m., training after work at 7 p.m.? That's fourteen hours awake — about 5.6% by the same rule of thumb (our arithmetic). It is an average across studies, not a measurement of your own session.
This is the line worth screenshotting for the group chat: after an early alarm, hours awake times 0.4% gives the rough size the review would anticipate. After an early alarm or a night without sleep, the later you push your session, the more it costs. Which raises a question nobody in the 5 a.m. crowd really wants to hear.
The 5 a.m. Alarm Is the Wrong Sacrifice
The researchers compared two kinds of short night, and the difference between them is brutal for gym culture.
Going to bed late but waking at the usual time produced no clear loss overall: the result was too scattered across studies to call, and skill and endurance tasks still fell. Waking up early after a normal bedtime produced a clear, consistent loss of around 7.4%.
Sit with that for a second. The sacrifice the fitness internet worships — the pre-dawn alarm, the "I'll just get less sleep and get it done" trade — is one of the two kinds of bad night the data punishes most; the other is no sleep at all. Sleeping through and stealing the lost hours off the front of the night is the one it mostly forgives, skill tasks aside.
The researchers drew the lesson themselves: when sleep loss is unavoidable and you have some say over timing, a late bedtime "would appear preferable" to an early alarm. When sleep has to be short, losing it off the front of the night cost less than losing it off the back. There's a second timing lever sitting right next to it.
Here's a wrinkle almost nobody mentions: not all strength is equal under sleep loss. Lower-body strength took a small but clear hit (about 3.4%), while upper-body strength barely budged. Your bench press is a safer bet on a bad night than your squat.
If You Train Anyway, Mornings Hold Up Better
The clock matters a second way — not just when you woke up, but when you train.
Across all the categories combined, morning sessions held up better than afternoon and evening ones. After a bad night, morning performance came in around 5% down. The afternoon and evening sat closer to 8%. Endurance went the other way, and only 8 of the 227 results came from tests after 6 p.m.
Put the two timing findings together and you have the authors' own advice. When a short night is coming and you have any say in it, the research leans toward going to bed late rather than getting up early. It also leans toward training earlier in the day, rather than dragging a depleted body into a later session.
The authors write it into their conclusion: if sleep loss is unavoidable, mimic a late bedtime rather than no sleep or an early alarm, and train in the morning. Of course, an average is only ever an average.
What the Average Can't Tell You
There's an honest catch in all of this, and it makes the findings more useful, not less.
Results were all over the map from study to study. That 7.5% is the middle of an enormous spread — the studies varied far more than chance alone would explain. The review pooled studies, not individual people, so it cannot tell you your own cost. The average hands you the direction and the rough size.
The review did not test why. Its introduction lists possibilities from earlier studies, among them lower motivation and a workout that feels harder; the authors left the mechanisms for others to explore.
And one limit deserves to be said plainly: nearly nine in ten participants across this research were men. Whether women pay the same tax is, honestly, still an open question. There simply wasn't enough data from women to answer it.
The Next Bad Night Is a Math Problem
So the next time the alarm goes off after four hours of sleep, you're not stuck guessing.
You've got a rough hierarchy: single max efforts mostly survive, while skill and precision work take the real hit. You've got a rule of thumb: after a night without sleep or an early alarm, every hour awake adds about 0.4% to the tab. And you've got a timing read: lose sleep off the front of the night, not the back, when you can; train earlier when you can't.
That's three tools the guy bragging about a three-hour night doesn't have. A bad night stops being a question of willpower and turns into something closer to a math problem — one you can actually see coming.
Which leaves the bigger question hanging. If one short night does this much to a single workout, what is it quietly doing to everything that workout was supposed to build?
You already know the split, so the call on a bad-sleep morning isn't whether to train — it's what to train.
If today was going to be technique work, skill practice, or anything fiddly and precise, that's the session a short night quietly sabotages. A day of heavy singles is the one working with what holds up; sets of repeated reps fell more, though that result is thin.
And if you've got any say in the clock, earlier beats later: after a night without sleep or an early alarm, the bill grows the longer you've been awake. That is close to the authors' own advice.
What this means for you
If your training leans on precision — shooting, serving, passing, ball control — this is the study that should give you pause. Skill-based tasks were the single most damaged category, dropping by roughly a fifth after a short night.
The researchers point to the heavy thinking load these tasks carry: sleep loss has been shown to dull reaction time, attention and decision making, so tasks that are also cognitively demanding are likely to be hit hardest. The takeaway the data offers is simple — a bad night costs you more than it costs the person next to you grinding out heavy singles.
If your whole game is the heavy single, this is the most reassuring result on the page. Maximal strength was the least-affected category, slipping under 3% — and it was also the most consistent finding across studies.
There's a deeper layer too: when the researchers broke strength down by the kind of short night, only a full all-nighter produced a clear drop. After merely short nights, the drop did not reach significance; after an early alarm it came close (about 4.5%). So the gym-floor instinct that "I can still hit my numbers tired" turns out to be closest to true for this exact group.
If you set an early alarm to guarantee the workout happens, the study has an uncomfortable note for you. Waking up early after a normal bedtime was one of the two kinds of bad night that hurt performance most clearly; the other was no sleep at all.
Going to bed late while keeping the usual wake time, by contrast, showed no clear loss overall, skill tasks aside. Read together, the numbers lean toward protecting the morning of sleep rather than sacrificing it — the opposite of what the pre-dawn alarm is built to do.
If you're a woman, the honest answer is that this study can't fully speak to you. Nearly nine in ten participants across all this research were men.
The researchers flagged this directly: there simply wasn't enough data from women to tell whether the same percentages hold. The smart read isn't to assume the averages transfer — it's to treat your own response over a few bad nights as better evidence than a number drawn mostly from men.
Before you change anything
This research is built almost entirely on healthy adults, and almost entirely on men — close to nine in ten participants. Everyone was an adult with no medical conditions, and every test followed a single short night, not a long run of them. If you're a woman, you were barely represented (about one in ten participants, too few to test separately); if you're carrying weeks of piled-up sleep debt, you're outside what these studies measured.
Almost all of these studies measured how long people slept, not how well they slept — so a night of poor-quality sleep at a normal length might behave differently. There's another gap: very few of the tests happened late in the evening, so the picture of nighttime performance is thin. And the data sorts exercise into neat categories, while a real sport like soccer blends several at once.
The headline number is solid — it held steady no matter which studies were pulled out of the pile. But two honest cautions come with it. First, results varied enormously from study to study, so the average cannot tell you what your own bad night will cost. Second, the finding that going to bed late "didn't matter" is better read as not yet proven, rather than proven safe — the result was simply too scattered to call.
Notice what just happened: a bad night stopped being a vague excuse and turned into a number you can see coming. But that number only covers the workout itself. The same short nights are also working on the body the workout is supposed to change — how much muscle protein your body builds after a night without sleep, how hard your appetite pushes back the next afternoon, and what your hormones are doing while you sleep too little. Each of those turns out to carry a price tag of its own.
What This Study Found
All findings from this paper, in plain language.
- A bad night drags down every kind of exercise, costing about 7.5% on average across all of them.
- Skill and accuracy tasks like shooting, serving, and throwing take the hardest hit, dropping around 21%.
- Raw maximal strength holds up best, slipping under 3% even after short sleep.
- Going to bed late but waking at the usual time showed no clear loss overall, except in skill and endurance tasks.
- After a night without sleep or an early alarm, performance drops about 0.4% for every extra hour spent awake.
- Workouts done in the morning held up better than afternoon and evening workouts after sleep loss.
- Lower-body strength took a small but clear hit while upper-body strength was barely affected.
- Muscular endurance showed a big drop, but the result was shaky and based on only a few studies.
- For strength, only a full all-nighter caused a clear drop — merely short nights mostly didn't.
- Endurance fell about 5.5%, and every kind of short night hurt it, though the late-bedtime and early-alarm results rest on two and three outcomes.
- When sleep must be short, losing it by staying up late beats waking up early for performance.
- The studies behind these numbers were generally good quality by the usual research checks.