Why does your weight change so much day to day?

Everything. A bathroom scale cannot separate fat from muscle, water, the food you ate last night or the coffee you just drank. It reports one number for the whole system, and on any given morning most of the movement in that number comes from the parts that are not fat.

The arithmetic settles this on its own. Gaining a kilogram of body fat takes a surplus of roughly 7,700 kcal, and a pound takes about 3,500. To gain a genuine kilogram of fat between Tuesday and Wednesday you would have to eat around 7,700 kcal above what you burned, in one day, on top of everything you normally eat.

A heavy meal out might put you 1,000 to 1,500 kcal above maintenance. That is real, and repeated often it matters, but as fat it is a couple of hundred grams at most. If the scale says you are 1.5 kg heavier the next morning, the other 1.3 kg is water, food and salt. It leaves again over the following days without you doing anything.

The same logic runs in reverse, which is the part people like less. A 2 kg drop in three days is not 2 kg of fat either. See how big a calorie deficit you actually need for what a realistic rate looks like.

Your body stores carbohydrate as glycogen, and glycogen holds water alongside it. Olsson and Saltin demonstrated in 1970 that total body water rises and falls with muscle glycogen, which is why a carbohydrate-heavy day shows up on the scale the next morning and why the first week of a low-carb diet produces a dramatic drop that then stops.

That early drop is frequently mistaken for rapid fat loss, and the plateau that follows is mistaken for failure. Neither reading is right. Both are glycogen and its water, moving out and then settling.

Salt works similarly over a shorter window. A high-sodium meal draws water into circulation and holds it there for a day or two. Restaurant food is usually far saltier than home cooking, which is a large part of why a meal out reads so heavily the next day.

Yes, and it has been measured. A 2014 study tracked daily weights in 80 adults across 4,657 measurements and found a consistent group-level rhythm: weight rises across the weekend, peaks on Sunday or Monday, and falls again from Tuesday onward.

The authors were explicit that this pattern should be treated as normal rather than as evidence of weight gain. More usefully, the people who showed the strongest weekday compensation were the ones most likely to lose or maintain weight over the long run. The rhythm is not a problem to eliminate; it is what ordinary eating looks like on a scale.

The practical consequence is simple. Comparing a Monday weight against a Friday weight will flatter you, and comparing Friday against Monday will alarm you. Neither comparison means anything. Compare like with like.

For most healthy adults, 1 to 2 kg (roughly 2 to 4 lb) between mornings is unremarkable. Larger swings are common after a big meal, a salty one, hard training, a long flight, poor sleep, or around menstruation, where hormonal water retention can add a couple of kilograms for several days and then release.

Bigger bodies tend to show bigger absolute swings, simply because there is more water in the system to move. If your daily variation looks large in kilograms, check it as a percentage of your body weight before concluding anything is unusual.

What is worth attention is a sudden, sustained gain of several kilograms over a few days accompanied by swelling or breathlessness. That is a reason to speak to a doctor rather than to adjust your calories.

You cannot remove the noise, but you can stop it moving around. Four rules do most of the work:

1. Same time, same conditions. First thing in the morning, after using the bathroom, before eating or drinking, wearing the same amount.
2. Weigh often, not rarely. Daily weighing feels worse but produces a better average. Weighing once a week means one noisy reading stands in for the whole week.
3. Read the weekly average, never a single morning. Compare this week's average against last week's. Any app that shows a trend line is doing exactly this.
4. Give it two to three weeks. That is roughly the shortest window in which a real trend separates itself from ordinary variation.

When the weekly average moves in one direction for two or three weeks running. That is the signal. Everything shorter is noise, and treating noise as signal is how people end up cutting calories further in a week where nothing was wrong.

If the average genuinely is not moving after three weeks of consistent eating, then your real maintenance level is lower than the estimate you started from. Adjust by 100 to 200 kcal and give it another few weeks. You can re-run the numbers in the Calorie Calculator, which shows every step of the arithmetic so you can see exactly which input moved your target.