Mifflin-St Jeor vs Harris-Benedict vs Katch-McArdle: Which to Use

The short answer

Use Mifflin-St Jeor unless you have a body fat percentage you really trust. A 2005 systematic review of the common equations found it predicted resting metabolic rate within 10% of the measured value in more people, lean and obese, than any other equation, and with the narrowest error range. Harris-Benedict is older and usually lands a little higher. Katch-McArdle ignores age and sex and uses lean body mass instead, so it can suit very lean or muscular people, but only when the body fat number going into it is good.

The BMR calculator shows all three side by side. Below is where each one comes from and what the gap between them means for you.

The three formulas

All three give calories per day. W is weight in kg, H is height in cm, A is age in years.

Formula	Men	Women
Mifflin-St Jeor (1990)	10W + 6.25H − 5A + 5	10W + 6.25H − 5A − 161
Harris-Benedict, revised (1984)	88.362 + 13.397W + 4.799H − 5.677A	447.593 + 9.247W + 3.098H − 4.330A
Katch-McArdle	370 + 21.6 × lean mass	same

Lean mass is weight × (1 − body fat %). A 200 lb man at 25% body fat has 150 lb, or 68.04 kg, of lean mass.

Where each formula comes from

Mifflin-St Jeor

In 1990, Mifflin, St Jeor and colleagues measured resting energy by indirect calorimetry in 498 healthy adults: 247 women and 251 men, aged 19 to 78, of whom 264 were normal weight and 234 had obesity. They fitted the equation above to that group. In the same paper they found that the 1919 Harris-Benedict equations overestimated measured resting energy by about 5%.

The paper also found that fat-free mass was the single best predictor of resting energy, better than weight. That is the idea Katch-McArdle is built on.

Harris-Benedict

The original equations date from 1918 and 1919. Most calculators today use the version re-fitted by Roza and Shizgal in 1984. They pooled the original 239 subjects with 98 more from Benedict’s later work and found the equations estimate resting energy in a normal subject with a precision of about ±14%. They also found the equations unreliable in malnourished patients. The coefficients in the table above are the 1984 ones, as listed in the Harris-Benedict equation summary.

Katch-McArdle

This one is 370 + 21.6 × lean body mass in kg. It is taught in exercise physiology, for example in this physiology lab text, which also notes you need a body fat estimate before you can use it. We could not find an original validation study for it, so we treat it as a cross-check rather than the headline number.

Three people, three formulas

Here is what each formula gives for three different adults. Numbers are rounded to the nearest calorie and match what the BMR calculator shows.

Person	Mifflin-St Jeor	Harris-Benedict	Katch-McArdle
Man, 35, 6 ft, 200 lb, 25% body fat	1,880	1,983	1,840
Woman, 60, 5 ft 3 in, 140 lb, 35% body fat	1,174	1,271	1,262
Man, 25, 5 ft 9 in, 180 lb, 10% body fat	1,792	1,881	1,957

The working for the first person: 200 lb is 90.72 kg and 6 ft is 182.88 cm.

• Mifflin-St Jeor: 907.2 + 1,143.0 − 175 + 5 = 1,880.
• Harris-Benedict: 88.4 + 1,215.4 + 877.6 − 198.7 = 1,983.
• Katch-McArdle: 370 + 21.6 × 68.04 = 1,840.

Three things stand out:

1. Harris-Benedict runs higher than Mifflin-St Jeor in all three cases, by about 90 to 105 calories here. That fits the 5% overestimate Mifflin’s team measured.
2. Katch-McArdle moves with body composition. It sits below the others for the man at 25% fat and above both for the lean 25-year-old, because at 10% body fat most of his 180 lb is lean tissue.
3. The spread is the real lesson. For the woman, the formulas are about 100 calories apart. Any one of them could be the closest for her. Nobody can tell from the numbers alone.

Which formula should you use?

Your situation	Best starting point	Why
Most adults	Mifflin-St Jeor	Most often within 10% of measured in the 2005 review
You know your body fat from a reliable scan	Compare Mifflin-St Jeor with Katch-McArdle	Lean mass drives resting energy
Very muscular or very lean	Katch-McArdle, as a cross-check	Weight-based formulas can’t see muscle
Older adult or from a US ethnic minority group	Mifflin-St Jeor, held loosely	The review found these groups underrepresented in the studies
Malnourished or acutely ill	None of these	Ask your care team; Roza and Shizgal found Harris-Benedict unreliable in malnutrition

Katch-McArdle is only as good as your body fat number

Because the formula multiplies lean mass by 21.6, every 10 points of error in body fat moves the answer a lot. Take the 200 lb man again:

• At 20% body fat: lean mass 72.57 kg, BMR 1,938.
• At 25%: lean mass 68.04 kg, BMR 1,840.
• At 30%: lean mass 63.50 kg, BMR 1,742.

A 10-point miss shifts his BMR by about 196 calories a day, double the gap between Mifflin-St Jeor and Harris-Benedict. If your body fat comes from a home scale or a quick tape estimate, Mifflin-St Jeor is usually the safer choice. The body fat calculator gives a Navy tape-measure estimate and the Army waist-to-height check, which are fine for tracking your own trend but still estimates.

What the review says about accuracy

The 2005 review is worth reading in its own words. It concluded that Mifflin-St Jeor is more likely than the others to estimate resting metabolic rate within 10% of measured, but that “noteworthy errors and limitations exist” when any equation is applied to one person. It also said the only way to remove the error is to measure resting energy with indirect calorimetry. Some hospitals and clinics offer that test.

So in practice:

• Pick a formula and stick with it, so your numbers are consistent over time.
• Turn BMR into daily calories with the TDEE calculator, which multiplies all three formulas by the same activity factor.
• Eat at that level for two to three weeks and watch your weight. If it holds steady, that intake is close to your real maintenance, whatever the formula said.

Common mistakes

• Treating BMR as daily needs. BMR is calories at complete rest. Your daily total is higher once you add movement.
• Mixing units. All three formulas use kilograms and centimeters. Using pounds and inches directly gives nonsense.
• Mixing up Harris-Benedict versions. The 1919 and 1984 coefficients differ. If a chart’s numbers don’t match the table above, check which version it uses.
• Chasing the highest number. If one formula gives you more calories, that doesn’t make it right for you.

To turn your result into a target for losing, keeping or gaining weight, use the calorie calculator.

These are population estimates, not a medical test. If you have a thyroid condition, are ill, pregnant or breastfeeding, or are under 18, talk to your doctor or a registered dietitian about your energy needs.