A pound of body fat stores roughly 3,500 calories, and a kilogram roughly 7,700. Strictly, asking how many calories is a pound has no single answer — it depends what the pound is made of — but for fat, 3,500 is the figure everyone means. That single number underpins almost every weight-loss rule of thumb you have heard, including the familiar advice to cut 500 calories a day to lose a pound a week. It is a genuinely useful planning figure. It is also routinely misapplied in three specific ways: to muscle, to scale weight, and to long stretches of time. This guide shows where the number comes from, does the arithmetic in the open, and explains exactly where it stops being true.
Where the 3,500 figure comes from
The number is not a measurement of a pound of tissue in a laboratory. It is a calculation, published in 1958 by Max Wishnofsky in the American Journal of Clinical Nutrition, and it rests on three inputs you can check yourself:
- A pound is 453.6 grams.
- Body-fat tissue is not pure fat — it is roughly 87% fat, with the remainder water, protein and cell structure.
- Fat carries about 9 calories per gram, the standard Atwater value.
Multiply them together:
- 453.6 g × 0.87 = 394.6 g of actual fat in a pound of fat tissue
- 394.6 g × 9 kcal/g ≈ 3,552 kcal
Rounded, that is the 3,500 everyone quotes. (The 9 kcal/g figure is an Atwater factor; see what kcal means for where it comes from.) Run the same sum in metric — 1,000 g × 0.87 × 9 — and you get about 7,830 calories per kilogram, which is why the metric figure is usually given as roughly 7,700. The two routes differ by a couple of percent because of rounding, not because of any disagreement about the biology.
Worth noticing: every input is an approximation. The 87% figure varies between people and between fat depots, and the 9 calories per gram is itself a rounded average. The output inherits all of that uncertainty, so treat 3,500 as a reasonable central estimate rather than a precise constant.
A pound of muscle is nothing like a pound of fat
This is the single most common misuse of the number, and it is worth doing the arithmetic rather than asserting it. Muscle is mostly water — roughly 70-75% — and water carries no energy whatsoever. The energy-bearing part is protein, which makes up about 20-22% of muscle by mass and carries 4 calories per gram, not 9.
Applying the identical method:
- 453.6 g × 0.22 = 100 g of protein in a pound of muscle
- 100 g × 4 kcal/g ≈ 400 kcal
| One pound of… | Energy-bearing content | Approximate energy |
|---|---|---|
| Body fat | ~87% fat at 9 kcal/g | ~3,500 kcal |
| Muscle | ~20-22% protein at 4 kcal/g | ~400 kcal |
| Water / glycogen weight | Water carries none | Effectively 0 |
Derived from standard tissue-composition figures and Atwater energy values. Ranges in the source figures put muscle somewhere around 360-450 kcal per pound; the point is the order of magnitude, not the decimal.
A pound of muscle therefore holds something like a ninth of the energy of a pound of fat. Two practical consequences follow. Building a pound of muscle does not require eating a 3,500-calorie surplus — the tissue itself is cheap in energy terms, and the real constraint is the training stimulus, not the calories. And losing a pound of lean tissue does not mean you were in a 3,500-calorie deficit; it usually means the deficit was too steep or protein was too low. If you want to track lean mass rather than scale weight, the lean body mass calculator and body fat calculator are the tools for that.
A pound on the scale is not a pound of fat
The 3,500 figure describes fat tissue specifically. The scale does not measure fat tissue — it measures everything, including a substantial and highly variable amount of water.
This is why the first week of a new diet is so misleading. Cutting carbohydrates depletes stored glycogen, and glycogen is stored with water, roughly 3 grams of water for every gram of glycogen. Losing a few hundred grams of glycogen therefore drops well over a kilogram from the scale in a matter of days, almost none of which is fat. The same mechanism runs in reverse when you eat normally again, which is the entire explanation for the “I gained four pounds in a weekend” experience. You did not gain four pounds of fat — at 3,500 calories a pound, that would have required eating 14,000 calories above maintenance.
That last calculation is actually the most useful way to use the number: as a sanity check on scale movements. Any weight change faster than the arithmetic allows is water, glycogen, or gut contents, not fat.
Why the rule overpromises over time
The arithmetic that gets people into trouble is the extrapolation. If a pound is 3,500 calories, then a 500-calorie daily deficit is 3,500 a week, which is a pound a week, which is 52 pounds a year. Almost nobody experiences that, and the reason is not willpower.
The rule quietly assumes your energy expenditure stays fixed while your body shrinks. It does not. A lighter body burns less at rest and costs less to move, so an unchanged intake represents a steadily smaller deficit. Kevin Hall and Carson Chow addressed this directly in a paper titled, bluntly, “Why is the 3,500 kcal per pound weight loss rule wrong?” (International Journal of Obesity). Their dynamic modelling shows the static rule substantially overpredicts long-run loss — where the simple sum forecasts around 52 pounds in a year, a realistic expectation is closer to half of that, with the remainder arriving only over subsequent years as the body settles at a new weight.
None of this means a deficit does not work. It means the timeline is slower and flatter than the multiplication implies. Set the deficit with the calorie deficit calculator, then check the pace against the weight-loss timeline calculator, which turns a weekly rate into a goal date and the milestone dates along the way.
How to actually use the number
Treat 3,500 calories per pound as a unit conversion, not a forecast. It is reliable for the things it was derived for and unreliable the further you stretch it:
- Good use — short-range planning. Translating a daily deficit into a rough weekly rate over the next few weeks is exactly what the figure is for.
- Good use — a reality check. Did the scale move faster than roughly a pound per 3,500-calorie deficit? Then it was not fat.
- Poor use — long projections. Anything beyond a couple of months needs a model that lets expenditure fall as weight falls.
- Wrong use — muscle. Use ~400 calories per pound, not 3,500.
- Wrong use — scale weight. The number describes fat tissue, and the scale reports everything.
A steady loss of about 0.5-1 kg (1-2 lb) per week is the range usually described as sustainable for healthy adults, and getting there depends far more on a deficit you can hold than on the precision of any conversion factor. Find your maintenance level first with the TDEE calculator; every number downstream depends on it.
This is general information, not medical advice. Energy needs and safe rates of weight change vary with age, health status, medication and pregnancy. If you have a medical condition, are pregnant or breastfeeding, or are planning a substantial change in weight, speak with a qualified healthcare professional or registered dietitian.
Sources: Wishnofsky M, “Caloric equivalents of gained or lost weight,” American Journal of Clinical Nutrition (1958); Hall KD & Chow CC, “Why is the 3500 kcal per pound weight loss rule wrong?” International Journal of Obesity (2013); Institute of Medicine (IOM/NASEM), Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat and Protein for Atwater energy values; CDC, Losing Weight and NHS, Healthy Weight for sustainable rate-of-loss guidance.