Harris-Benedict Calculator

What the Harris-Benedict equation predicts

The Harris-Benedict equation predicts basal metabolic rate, the energy a body spends at complete rest on breathing, circulation, cell repair and holding its temperature steady. James Arthur Harris and Francis G. Benedict published it in 1918 from measurements of 136 men and 103 women at the Nutrition Laboratory of the Carnegie Institution of Washington, and the full paper is available through PubMed Central. Their regression weighs weight in kilograms, height in centimeters and age in years, then adds a constant for sex.

The result is a 24 hour figure in kilocalories, the same calorie printed on food labels. It is not a target for how much to eat. Rest covers only part of a day, and meals, walking, work and training all sit on top of it. When the equation was reevaluated in 1984, the authors reported that it predicts resting energy expenditure in normally nourished adults with a precision of about 14 percent, so treat any single output as a bracket rather than a fact.

The 1918 report came out of a measurement program at the Carnegie Institution's Nutrition Laboratory, not a survey. Harris and Benedict worked from 239 subjects and published look-up tables for weights from 25.0 to 124.9 kg, statures from 151 to 200 cm and ages from 21 to 70. A modern user outside those ranges is extrapolating, and nothing on the screen warns that it happened.

Two things follow from that definition and explain most of the confusion around the equation. The first is that the output is a prediction, not a measurement. Indirect calorimetry with a metabolic cart is the measurement, and it costs far more than a formula. The second is that the equation predicts resting burn, so an 1,837 calorie output says nothing about a day that includes a commute, a gym session and a dog walk. The arithmetic below carries both facts through every example.

The original 1918 equations

Two equations carry the 1918 name, one for each sex. Weight is in kilograms, height in centimeters and age in years:

Every term earns its place. Weight carries the most: 13.7516 calories for each kilogram for a man and 9.5634 for a woman. Height adds a smaller amount, and age subtracts: a man loses 6.755 calories from his resting rate each birthday, a woman 4.6756.

Metric units are not optional inside the formula. Pounds divide by 2.20462 and inches multiply by 2.54, and a value entered in the wrong unit returns a number that looks plausible and is off by hundreds of calories.

The 1984 revision by Roza and Shizgal

Sixty six years later, A. M. Roza and H. M. Shizgal refit the equation against a newer sample of 337 adults, 168 men and 169 women, and published the revision in the American Journal of Clinical Nutrition (PMID 6741850). The revision keeps the same four term shape and changes every coefficient.

The direction of each change tells you where the versions drift apart. The revision eases the age penalty for men by 1.08 calories a year and nearly doubles the height term for women by 1.25 calories a centimeter, so the gap between the equations grows with age for men and with height for women. On the two examples below, the 1918 form still lands 5 calories higher for the man and 16 higher for the woman. Both versions stay in service because textbooks, diet software and fitness trackers were built on them, and neither is wrong about its own sample.

Worked example: a 35 year old man, 180 pounds, 5 feet 10 inches

Conversions first: 180 lb divided by 2.20462 is 81.65 kg, and 5 feet 10 inches is 70 inches, so 70 x 2.54 = 177.8 cm.

Now the 1984 revision term by term:

The total is 1,837 calories a day at rest. Running the same body through the 1918 form gives 1,842, a gap of 5 calories. Ten years later at the same weight and height, the age term grows by 5.677 x 10 = 56.77 and the total falls to about 1,780.

Activity sits on top. At a moderate multiplier of 1.55, his daily estimate is 1,837 x 1.55 = 2,847 calories, which is the number a maintenance diet should be built from, not the 1,837.

Worked example: a 28 year old woman, 145 pounds, 5 feet 4 inches

Conversions: 145 lb is 65.77 kg and 5 feet 4 inches is 162.56 cm. The 1984 revision gives 447.593 + 9.247 x 65.77 + 3.098 x 162.56 - 4.330 x 28 = 1,438 calories a day.

The 1918 form gives 1,454 for the same woman, and the gap is wider here than for the man, mostly because the revision nearly doubled the height term for women. At a light activity multiplier of 1.375, her daily estimate is 1,438 x 1.375 = 1,977 calories.

Swap in the 1918 constants and her resting figure rises to 1,454; at the same multiplier the daily estimate is 1,454 x 1.375 = 1,999, a difference of 22 calories a day, which compounds to about 8,000 calories over a year of tracking.

That scale is why the two versions are usually quoted together. A gap of 5 to 16 calories is small enough to ignore for a week of eating and large enough to show up across a year of tracking, which is the scale on which a resting estimate matters.

Pounds, inches and the metric trap

The equation wants kilograms and centimeters. The conversions are fixed, so a short reference table saves the arithmetic:

The trap is entering a pound figure where a kilogram belongs. Type 180 into the weight box of a metric calculator and the weight term becomes 13.397 x 180 instead of 13.397 x 81.65, which is 1,318 calories too high. That single slip pushes most results past any honest error band, and it is the most common reason two calculators disagree.

Small rounding is more forgiving. The same man entered as 82 kg and 178 cm instead of 81.65 and 177.8 lands at 1,842 instead of 1,837, a spread of about 6 calories that no diet can feel.

From resting rate to daily calories

The equation stops at rest. Daily needs come from multiplying by an activity factor, and the factors below are the same conventions used across consumer calculators:

The spread from sedentary to athlete is 1,286 calories for the man, which dwarfs the 5 calorie gap between the two equations. Picking the multiplier honestly matters more than picking the formula.

Two public tools estimate needs from measured data if you want a second opinion: the NIDDK Body Weight Planner models weight change over time, and the USDA DRI calculator returns estimated energy requirements from the national reference equations.

Know the limits. The 1984 reevaluation reported a precision of about 14 percent, and the original tables ran from age 21 to 70, so children, pregnancy and serious illness are outside what the equation can see. For most people the practical move is to start here, hold intake steady for two weeks, and let the scale correct the estimate.

One more check is worth running before you trust a number: compare it with a second equation. The Harris-Benedict pair and Mifflin-St Jeor usually land within 100 calories of each other for people in the middle of the weight range, and a spread larger than that points to an input the formula cannot see, such as unusual lean mass or a body fat percentage far from average. The TDEE calculator carries the same idea further with goal calories, and the plain BMR calculator compares three equations side by side.

What changed after 1984

Harris-Benedict was never the final word, and the decades since 1984 added alternatives rather than replacements. The 1985 joint FAO/WHO/UNU expert consultation on energy and protein requirements adopted the Schofield equations for population level estimates, built from a wider international dataset. The 1990 Mifflin study argued that its sample reflected contemporary North Americans more closely than the 1918 cohort, and that equation became a common default inside dietetic software.

Clinical practice kept every option. A dietitian choosing between Harris-Benedict, Mifflin-St Jeor and Schofield is picking the sample that resembles the patient in front of them, and the 1984 paper's own caveat still applies: the arithmetic predicts normally nourished adults within about 14 percent and is unreliable in malnourished patients, who need measured expenditure instead.

Consumer devices took another path. Fitness watches and phone apps rarely disclose which equation they use, and the number moves when a vendor updates firmware. When a tracker and this page disagree by 100 calories, the difference is usually the equation plus an activity guess, not one side making an arithmetic mistake.