Mifflin-St Jeor vs Harris-Benedict: how the BMR equations differ

What the equations predict

Basal metabolic rate is the energy the body uses at complete rest. Measuring it needs a lab method such as indirect calorimetry, which works from the gases a person breathes in and out.

A prediction equation is a shortcut. Researchers measure a group, then use regression to find the weights for each input that best match the measurements. The Mifflin paper calls its result resting energy expenditure, or REE. In everyday use, both terms describe the same kind of estimate.

Because each equation is fitted to one sample, it works best for people who resemble that sample. The equations were fitted separately for men and women, which is why the ‘Sex’ input matters.

Harris-Benedict: 1918–1919, revised in 1984

Harris and Benedict published their equations in 1918 and 1919 through the Carnegie Institution. The data came from 239 people, 136 men and 103 women, aged 16 to 63, collected over about ten years.

In 1984 Roza and Shizgal refitted the same form of equation to a larger data set. It held 337 people, 168 men and 169 women, with a wider age range. The calculator uses this revised version.

Mifflin-St Jeor: 1990

M.D. Mifflin, S.T. St Jeor and colleagues published their equation in the American Journal of Clinical Nutrition in 1990. They measured 498 healthy adults aged 19 to 78, including both normal-weight and obese people, by indirect calorimetry.

Their regression on weight, height, age and sex explained about 71% of the variation in measured REE. Adding relative body weight or fat distribution did not improve it in a meaningful way. Fat-free mass was the strongest single predictor, but most people do not know theirs, so the published equation uses total weight instead.

The paper prints the coefficients as 9.99 for weight and 4.92 for age. The widely used form rounds them to 10 and 5, and that is what the calculator uses. For the man in the example below, the unrounded version gives 1,781.6 kcal, about 2 kcal more.

Why the results differ

The two equations disagree for three plain reasons. They were fitted to different people, in different decades, with different methods. In the Mifflin study, the Harris-Benedict equations overestimated measured energy expenditure by about 5%.

That finding is not universal. A 2023 study in the journal Metabolites tested both on its own sample and found lower error for Harris-Benedict than for Mifflin-St Jeor. Each test sample favors the equation that fits its people best.

The structure also differs. Harris-Benedict gives men and women separate slopes for weight, height and age. Mifflin-St Jeor uses one set of slopes and moves the whole line by a fixed 166 kcal between men and women.

Reading the two numbers side by side

The calculator shows both results at once (plus two lean-mass equations if you add ‘Body fat %’), and the ‘Formula’ setting picks which one leads. In the two examples above, the gap is 74 kcal for the man and 64 kcal for the woman.

Neither number is your true BMR. Each is the average for people in the original sample with your weight, height, age and sex. Individuals vary around that average, partly because two people of the same weight can carry very different amounts of muscle. The equations are for adults, and they are not made for pregnancy or medical conditions that change metabolism.