Calorie Calculator

The two numbers, and what separates them

The first figure is basal metabolic rate: what the body spends on the heart, lungs, kidneys and temperature with no activity at all. The second is total daily energy expenditure, which is the first number times an activity factor. Strictly the equations predict resting metabolic rate, a few percent above true basal rate because it is measured under looser conditions.

Every row of the goal table starts from the second number. The 10% and 20% adjustments apply to total expenditure, not to your basal rate.

Where the four equations came from

Mifflin-St Jeor (1990) came from 498 healthy adults aged 19 to 78, roughly half of them obese, with resting expenditure measured by indirect calorimetry. Frankenfield's 2005 review of 29 studies found it the most accurate of the common equations in healthy adults, and the Academy of Nutrition and Dietetics recommends it. Use it as the default.

Harris-Benedict is where labelling gets slippery. Three equations circulate under that name: the 1919 original, the 1919 data re-fitted with modern regression, and the 1984 Roza-Shizgal revision, which is the one this calculator runs. Almost no site says which. They agree in middle age and diverge at the extremes, because the 1919 version penalises age far harder in men.

The option labelled Katch-McArdle, 370 + 21.6 times lean body mass, has no paper of that name behind it: Cunningham proposed the coefficients in 1991 and Katch and McArdle reproduced them in a textbook. Its accuracy cannot exceed the accuracy of the body-fat figure you type in, and a 5-point error on a 65 kg person moves the estimate by about 70 kcal a day, an ordinary miss for a bioimpedance scale. With a DEXA figure it is the strongest option here: in recreational athletes aged 18 to 35 it reached 84.9% accuracy in men and 78.4% in women, while Harris-Benedict, Mifflin, WHO, Schofield and Owen were all below 50%.

Ganpule's National Institute of Health and Nutrition equation (2007) came from 137 Japanese adults and appears in Japan's official dietary reference intakes. It is offered on the Japanese and Korean versions of this page as the only choice drawn from an East Asian cohort.

The 1.2 problem

The values 1.2, 1.375, 1.55, 1.725 and 1.9 are widely called the Harris-Benedict activity multipliers. They are not. Harris and Benedict measured basal metabolism only, in post-absorptive subjects at complete muscular rest, and published no multipliers. The ladder is a fitness-industry convention usually credited to an exercise physiology textbook, and no peer-reviewed paper derives it.

More seriously, 1.2 does not describe a person who gets out of bed. The FAO/WHO/UNU report on human energy requirements puts the activity levels free-living adults can sustain long term at about 1.40 to 2.40, and offers a mean PAL near 1.21 for short-term survival of totally inactive dependent people in conditions of crisis. That report's own sedentary example, eight hours of sleep plus eight hours of seated office work and light domestic activity, works out at 1.53.

The sourced bands give you something to anchor on. FAO puts sedentary or light lifestyles at 1.40 to 1.69, and 1.70 to 1.99 for a desk worker who also does about an hour of moderate-to-vigorous exercise. Japan's health ministry states its adult levels in hours: 1.65, 2.06 and 2.53 hours a day at 3.0 to 5.9 METs, giving factors of 1.50, 1.75 and 2.00.

For the man above, with a predicted BMR of 1,568, choosing 1.2 gives 1,881 kcal and 1.5 gives 2,351. That 470 kcal gap is four times the entire spread across the four formulas.

The 3,500-calorie rule, and why it stops working

The weekly column takes your daily surplus or deficit, multiplies by seven and divides by 7,700 kcal per kilogram. Wishnofsky produced that constant in 1958 by assuming a pound of adipose tissue is about 87% fat, multiplying by roughly 9 kcal per gram, and rounding to 3,500 kcal per pound. Note what it describes: adipose tissue, which is fat plus water and cell structure, not pure fat and not body weight. The calculator applies it to body weight, which includes lean tissue and water too.

Hall's 2008 analysis showed the deficit required per kilogram lost rises with how much fat you started with. Roughly 7,700 kcal/kg fits people carrying more than about 30 kg of body fat; below about 25 kg it substantially overestimates the deficit needed, so leaner people initially lose faster per calorie cut than the rule predicts.

The bigger failure is that the rule is static. It assumes expenditure never changes, and it does: as you lose weight there is less of you to maintain and less to move, so the deficit shrinks toward zero. Real weight curves flatten; a straight line never does. The US National Institute of Diabetes and Digestive and Kidney Diseases dropped the rule for a dynamic simulation, the Body Weight Planner. The replacement from Hall and colleagues in the Lancet in 2011: a permanent change of about 24 kcal a day produces roughly 1 kg of eventual weight change, half within a year and 95% within three.

So read 0.44 kg a week as about 0.4 kg a week at first, tapering. Multiplying it by 52 will not describe your year.

How much precision is actually there

Roza and Shizgal put the accuracy of the Harris-Benedict equations at about 14% either way in normally nourished people, and unreliable in malnourished patients. Even the best-matched case leaves error: in 365 healthy Japanese adults with basal rate measured directly, the most accurate equation still carried a total error of 99 to 125 kcal a day. On a 2,000 kcal maintenance level, 10% is 200 kcal, more than a meal. The displayed digits do not change that, and both the whole-kcal figures and the two-decimal weekly weight sit far inside the noise.

Check the age range. Mifflin-St Jeor was derived on 19 to 78 year olds, the Japanese equation on 20 to 74 year olds with validity confirmed from 18 to 79. This calculator accepts 18 to 79 and returns nothing outside that, because outside it these equations give a confident number with no evidence behind it. Growing children need age- and sex-specific references instead.

Watch the combinations. The lowest activity factor plus the −20% row returns 1,505 kcal for the man above, below the 1,568 kcal basal rate printed two boxes higher. That is arithmetic, not a recommendation.

Using it in practice

Treat the maintenance figure as a starting hypothesis rather than a measurement. Eat at the estimate, weigh yourself on the same terms for two to four weeks, and let the trend correct the number: flat weight means the estimate was right for you, drift means adjust by 100 to 200 kcal and repeat. Judge it over weeks, not days, since day-to-day swings are water, glycogen and gut contents.

This page is an estimate for general reference, not medical advice and not a diagnosis. If you have a medical condition, are pregnant or breastfeeding, or plan a substantial change in body weight, agree the target with a doctor or registered dietitian.

Frequently asked questions

Which BMR formula is the most accurate?

Mifflin-St Jeor for general adults: Frankenfield's 2005 review of 29 studies found it the most accurate of the common equations, and the Academy of Nutrition and Dietetics recommends it. If your body-fat percentage comes from DEXA rather than a bathroom scale, the Cunningham equation labelled Katch-McArdle beats it, reaching 84.9% accuracy in men and 78.4% in women among recreational athletes aged 18 to 35.

Is Harris-Benedict the original formula?

Probably not the one you are running. Three equations circulate under that name, and this calculator uses the 1984 Roza-Shizgal revision. All versions read high, and highest in East Asian populations: against measured values in 365 Japanese adults, Harris-Benedict overestimated by 99 kcal a day in men and 150 kcal a day in women.

I have a desk job. Is that sedentary, 1.2?

No. FAO/WHO/UNU puts the sustainable floor for free-living adults at about 1.40, and its own example of an office worker with eight hours of seated work comes out at 1.53. A PAL near 1.21 is what that report reserves for totally inactive dependent people in a crisis. If you commute and sit at a desk, pick 1.375 and treat its output as a floor.

Will cutting 500 calories a day lose me a pound a week?

Roughly, for a few weeks. Then it stops, because expenditure falls as you get smaller and the real deficit shrinks. The NIH replaced the 3,500-calorie rule with a dynamic model for that reason. The modern rule of thumb: a permanent change of about 24 kcal a day produces about 1 kg of eventual weight change, half within a year.

Can I eat below my BMR to lose weight faster?

The lowest activity level plus the −20% goal returns 1,505 kcal for a man whose predicted basal rate is 1,568. That is what the arithmetic produces, not what the page recommends. Very-low-calorie intakes carry real risks of lean-mass loss and nutrient deficiency and belong under medical supervision.