How Much More You Eat After Poor Sleep

A few hundred calories on average — and a spread wide enough to swallow the average#

After a bad night's sleep you will probably eat more the next day, and the number most often quoted is a few hundred calories — controlled studies land somewhere between roughly +250 and +560 for a short night. That is the answer to the title, and if you stop there you'll draw exactly the wrong lesson from it. The average is real, but it's the mean of a distribution wide enough that "you'll eat 300 more calories tomorrow" is less a rule than a coin-toss. In the studies that actually report the spread, some people eat over a thousand extra calories and some eat less than they did on a full night's sleep.

So the useful framing isn't a penalty you owe after every short night, but a tilt in the odds — the deck reshuffled toward eating more, by an amount you can't know in advance. This spoke is the quantification; the broader case for sleep as a weight lever is the pillar, and how the surplus accumulates into weight over months is the neighbouring article on how sleep loss drives weight gain, which owns the pooled accounting. Here the question is narrower and more personal: after one bad night, how much more — and how sure can you be?

One short night, measured#

The tightest single-night experiment is small and clean. Twelve healthy young men slept either 8 hours or 4 hours for one night, in a randomized crossover so each man was his own control, then ate freely the next day¹. After the short night they consumed 559 ± 617 kcal more — about 22% more energy — and reported stronger hunger before both breakfast and dinner.

Read the ± before you spend the 559. The standard deviation, 617, is larger than the mean. In plain terms: the man-to-man variation was bigger than the average effect itself, which is the statistical signature of something real on average and wildly uneven in the individual. One bad night did not hand each of these men a 559-calorie bill; it handed the group an average of 559, assembled from some men who ate enormously more and some who barely moved.

The standard deviation was larger than the mean. One short night didn't add 559 calories to each man — it added an average of 559, built from some who ate vastly more and some who didn't budge.

A longer Mayo Clinic trial found the same shape at a bigger dose. Over eight nights of sleep cut to two-thirds of normal, the restricted group ate 559 kcal a day more (SD 706) while well-rested controls drifted down 118, for a net difference of 677 kcal/day (95% CI, 148–1,206)². Same headline number, same enormous scatter — an SD once again bigger than the mean — and, notably, no change in leptin or ghrelin, which is the hunger-hormone story coming up short again.

The average hides who you are#

If two studies with standard deviations bigger than their means aren't enough, one analysis went looking at the spread directly. Pooling crossover data, McNeil and St-Onge found individual changes in energy intake after sleep restriction ranging from −813 to +1,437 kcal a day, and concluded plainly that "not all participants were negatively impacted"³. That band is nearly 2,300 calories wide. Some people, on short sleep, eat like they skipped a meal; others eat like they added one and a half.

The obvious question is whether the split is predictable — and the tidy guesses don't hold. Sex did not predict the response; men and women scattered alike. The only factor that tracked the size of the effect was study phase order, an artefact of which condition someone did first, not a trait you could screen for. So there's no known way to look at yourself and say which kind of sleeper you are. You find out by watching what actually happens.

Study	Sleep dose	Average extra intake	The spread
Brondel, 2010	1 night, 4h vs 8h	+559 kcal (22%)	± 617 (SD > mean)
Calvin, 2013	8 nights, ~2/3 normal	+559 kcal/day	SD 706 (SD > mean)
McNeil & St-Onge, 2017	Crossover pool	—	−813 to +1,437 kcal/day

The pooled meta-analytic average across many such trials is about +385 kcal a day, and even that carries a confidence interval from 252 to 517⁴ — a range, not a point. That pooled figure and its late-night timing belong to the weight-gain article; the point here is that every credible version of this number arrives wearing an error bar.

You don't quietly eat less later to cancel it#

The tempting hope is that a body eats more one day and compensates the next — appetite self-correcting over the week. The short-run evidence doesn't support it. In the eight-day Mayo trial the surplus persisted across days rather than washing out, and expenditure didn't rise to meet it². Brondel's men did move a little more the afternoon after their short night, but nowhere near enough to offset several hundred extra calories eaten — and that small activity bump is the exception, not a reliable brake¹. Why there's no metabolic rebate on the burn side is the subject of does poor sleep slow your metabolism: tired bodies don't burn less, they mostly just eat more.

One more twist makes the surplus sneakier than its size suggests. It isn't spread evenly across better and worse foods — the extra calories skew toward calorie-dense, palatable choices, which is why it's chips and not carrots when you're tired. A few hundred calories of vegetables would be hard to eat by accident; a few hundred calories of the food a tired brain actually reaches for is two handfuls.

Playing the odds after a short night#

The practical move follows from the distribution. Because you can't know in advance whether tomorrow is a +50 day or a +900 day, the reliable strategy isn't willpower on the day — it's making the surplus visible so you can see which day you got. A single bad night isn't a debt to pay off with a punishing workout; it's a day when your odds of eating more are elevated by an unknown amount, and the only way to learn the amount is to keep logging through it instead of writing the day off.

That reframes the guilt, too. If the effect were a fixed 300-calorie tax, overeating on a tired day would look like a personal failure of the same size every time. It isn't fixed, it isn't predictable, and for some people on some nights it isn't there at all. Treat a short night as a raised probability, not a verdict — and let the record, not your memory, tell you how it actually landed.

FAQ#

Will everyone overeat after a bad night's sleep?#

No. On average people eat a few hundred calories more, but the individual response is a wide distribution: in pooled crossover data the change ran from −813 to +1,437 kcal a day, and "not all participants were negatively impacted"³. Sex didn't predict who overate. Some people genuinely eat the same or less — you find out by tracking, not by assuming.

Is the "+300 to +500 calories" figure something I'll definitely eat?#

Treat it as an average, not a personal quota. One clean single-night study measured +559 kcal — but with a standard deviation of 617, larger than the mean itself¹. A larger pooled estimate lands near +385 kcal/day with a 95% confidence interval of 252 to 517⁴. The number is a distribution with a wide error bar, not a fixed amount you can count on eating.

Does the extra eating balance out over the next few days?#

Not on the evidence available. Across eight days of restricted sleep the surplus persisted rather than self-correcting, and energy expenditure didn't rise to offset it². A small bump in afternoon activity has been seen¹, but it doesn't come close to cancelling several hundred extra calories. If you sleep badly for a run of nights, don't count on a spontaneous rebound.

Sources#

This article was researched and drafted with AI assistance and reviewed for accuracy by the BurnWeek team. It is general information, not medical advice. How we research and correct our articles →