The number is about right. The straight line drawn through it is not.#
As an estimate of the energy stored in a pound of human fat tissue, 3,500 calories is defensible. That part of the folklore survives scrutiny: modelling of the composition of lost weight found the rule of thumb "approximately matches the predicted energy density of lost weight in obese subjects with an initial body fat above 30 kg"2. So if someone asks how many calories are in a pound of fat, 3,500 is a reasonable answer and this article is not going to pretend otherwise.
The failure is what everyone does next. Take a number that describes a substance and use it as a schedule — 500 a day, a pound a week, fifty-two pounds a year — and you have built a straight line out of a quantity that was never linear. This is not a case where the evidence is split and reasonable people differ. A working group of energy-balance researchers stated flatly that "the 3500 kcal per pound rule should no longer be used," because it is routinely misapplied to imply a linear weight change that does not occur4. The arithmetic was never the problem. The extrapolation is. If you want the model that replaced it, that is the subject of the calorie deficit pillar; this page is about why the old one broke.
Where 3,500 came from#
It is worth knowing the rule's provenance, because it explains its shape. The figure traces to a 1958 paper by Max Wishnofsky in the American Journal of Clinical Nutrition, which reviewed the existing literature on body composition and concluded that the caloric equivalent of a pound of body weight gained or lost is 3,5001.
Notice what that is and is not. It is a careful synthesis of what adipose tissue contains — a substance measurement. It is not a trial of what happens to a human being who eats 500 fewer calories for a year, because no such trial informed it. Wishnofsky described a tissue. Six decades of diet advice turned his description into a promise about time, which is a thing he never claimed and the paper never tested.
You have never lost a pound of fat#
Here is the mechanism the straight line ignores: nobody loses pure adipose tissue. You lose a mixture — fat, glycogen, the water bound to that glycogen, and some lean tissue — and lean tissue is largely water, so it carries far less energy per gram than fat does. The mixture is not fixed. It depends on how much fat you had to begin with, and it changes as the diet goes on.
On the first point, Hall's model predicts that a larger cumulative deficit is needed per unit of weight lost in people with more initial body fat, and that the rule "overestimates the cumulative energy deficit required per unit weight loss for people with lower initial body fat"2. A leaner person shifts more of their loss to lean tissue, which is cheap in energy terms — so each pound off the scale costs them less than 3,500, not more.
On the second point, there are measurements. Twenty-three overweight adults in the CALERIE study were tracked by DEXA through weeks 4 to 24 of energy restriction, and the energy content of each kilogram of weight change was calculated directly3:
| Measurement | Energy content of lost weight |
|---|---|
| CALERIE, week 4 | 4,858 ± 388 kcal/kg |
| CALERIE, week 6 (and stable after) | 6,041 ± 376 kcal/kg (P < .01 vs week 4) |
| Kiel study, women (n = 75 total) | 6,804 ± 226 kcal/kg |
| Kiel study, men | 6,119 ± 240 kcal/kg (P < .05 vs women) |
| What the 3,500 kcal/lb rule implies | ~7,700 kcal/kg |
That last row is my own conversion, and it is the only line in the table that is not measured: Hall's paper gives the rule's metric form as 32.2 MJ per kilogram, which works out to roughly 7,700 kcal/kg. Every measured value sits below it — and the week-4 figure sits far below it, because early loss is disproportionately glycogen and water. By week 6 the mixture had shifted toward fat and the energy content had climbed by about 1,200 kcal/kg, then stopped moving. The composition of what you are losing is a moving target for the first month or more, which is also the difference between losing fat and losing weight.
The rule is wrong in both directions at once#
This is the part that makes the 3,500 rule genuinely destructive rather than merely imprecise, and it is why "it's roughly right" is such a poor defense of it.
Early on, the rule undersells you. If a kilogram of the weight you are shedding in week 4 contains 4,858 kcal rather than 7,700, then a given deficit strips off substantially more weight than the rule predicts. Your first fortnight beats the forecast. It feels like your metabolism is cooperating. It is mostly glycogen and its water leaving.
Later, the rule oversells you. Extend the straight line and it promises a constant weekly payout forever. Real physiology does not: expenditure falls as the body it powers gets smaller, so the curve flattens. In the dynamic model, the body-weight response to an intake change has a half-time of about a year, and a sustained change of 100 kJ/day is worth about 1 kg — eventually, not weekly5. Month six delivers a fraction of what the line drew.
The rule makes your second week look like a miracle and your sixth month look like a failure. Both readings are the same straight line, lying in opposite directions.
And it is the second lie that does the damage, because it arrives with a moral attached. A dieter who was promised a pound a week and got half of it in month five does not conclude that the arithmetic was naive. They conclude that they cheated, or that their metabolism is broken. Neither is implied by the data. The prediction was.
What replaced it: a curve, not a constant#
The replacement is not another constant — it is a model that lets expenditure move. The NIH's dynamic model of adult metabolism simulates that adaptation and drives the agency's own Body Weight Planner; it predicts a decelerating curve toward a new steady weight rather than an endless slope, with roughly 50% of the eventual change arriving in a year and 95% in about three (Hall et al., 2011; Hall et al., 2012).
The practical upgrade is smaller than it sounds. Stop treating any calorie figure as a promise with a date on it. Your intake number carries real error — the layers are audited in how accurate calorie counting really is — your maintenance number is a population estimate wearing your name (TDEE explained), and now the conversion rate between calories and kilograms turns out to be a variable too, moving with your body fat, your sex, and how many weeks in you are. Three estimates multiplied together do not make a schedule. They make a direction, which is all a deficit was ever going to give you, and the reason calorie counts are worth reading as ranges rather than facts.
FAQ#
How many calories are in a pound of fat?#
About 3,500, and that figure holds up reasonably well as a description of pure adipose tissue — modelling found it approximately matches the energy density of weight lost by people carrying more than 30 kg of body fat2. The catch is that a pound off the scale is never a pound of pure fat, so the conversion rate that matters to a dieter is lower and keeps moving.
Why doesn't cutting 500 calories a day lose a pound a week?#
Because two things change underneath you. The mixture you are losing shifts — early weight is largely glycogen and water, measured at 4,858 kcal/kg in week 4 versus 6,041 by week 63 — and your expenditure falls as your body gets smaller, so the same 500 kcal buys less each month. The result is a decelerating curve, not a line.
Is the 3,500-calorie rule completely useless?#
Not completely, and it is worth being precise about the failure. As a rough energy content for fat tissue, it is fine. As a one-week sanity check on a plan, it is survivable. As a projection over months, it is the source of more unearned guilt than any other number in dieting, which is why a research panel recommended retiring it outright rather than qualifying it4.
Sources#
- Wishnofsky M. Caloric equivalents of gained or lost weight. Am J Clin Nutr. 1958;6(5):542-6.
- Hall KD. What is the required energy deficit per unit weight loss? Int J Obes (Lond). 2008;32(3):573-6.
- Heymsfield SB, Thomas D, Martin CK, Redman LM, Strauss B, Bosy-Westphal A, Müller MJ, Shen W, Nguyen AM. Energy content of weight loss: kinetic features during voluntary caloric restriction. Metabolism. 2012;61(7):937-43.
- Hall KD, Heymsfield SB, Kemnitz JW, Klein S, Schoeller DA, Speakman JR. Energy balance and its components: implications for body weight regulation. Am J Clin Nutr. 2012;95(4):989-94.
- Hall KD, Sacks G, Chandramohan D, Chow CC, Wang YC, Gortmaker SL, Swinburn BA. Quantification of the effect of energy imbalance on bodyweight. Lancet. 2011;378(9793):826-37.
- NIDDK. Body Weight Planner. National Institute of Diabetes and Digestive and Kidney Diseases.
