Skip to main content
aifithub
weight loss Comparison

Mifflin-St Jeor vs Harris-Benedict: Which Is More Accurate

If you've used more than one TDEE calculator, you've probably noticed they give different numbers. The reason: they use different BMR formulas under the hood. The two most common are Mifflin-St Jeor (1990) and Harris-Benedict (revised by Roza & Shizgal, 1984). The difference typically ranges from 50-200 calories — small enough to seem insignificant, but large enough to make or break a tight deficit over weeks.

By AI Fit Hub · AI Fit Hub Team
Best Next MoveNutrition

TDEE Calculator

Estimate your daily energy expenditure with Mifflin-St Jeor + activity factors.

CalculatorOpen ->

On This Page

Education · Not medical advice. Output is deterministic math from your inputs.Editorial standardsSponsor disclosureCorrections

Mifflin-St Jeor (1990) Option

The newer, ADA-recommended formula. Uses weight, height, age, and sex. Validated against indirect calorimetry in multiple studies.

Pros

  • Predicted BMR within 10% for 82% of subjects in the ADA's 2005 comparative review
  • Better accuracy across both normal-weight and obese populations
  • Recommended by the American Dietetic Association as the standard for clinical use
  • Simpler formula with fewer coefficients

Cons

  • Doesn't account for body composition (lean mass vs fat mass)
  • May underestimate for very muscular individuals
  • Less accurate for underweight individuals (BMI < 18.5)

General population, anyone without a body fat percentage measurement, clinical and research settings

Harris-Benedict Revised (1984) Option

The older standard, revised in 1984 from the original 1919 formula. Uses the same inputs. Tends to estimate ~5% higher than Mifflin-St Jeor.

Pros

  • Long track record — the original (1919) is one of the most cited equations in nutrition science
  • 1984 revision improved accuracy over the original
  • May be more accurate for underweight individuals

Cons

  • Overestimates BMR by ~5% on average compared to measured values
  • Only predicted BMR within 10% for 69% of subjects in the ADA review
  • Higher error rate in obese individuals — overestimates more as body fat increases
  • No longer the ADA's recommended formula

Historical comparison, checking if your Mifflin-St Jeor result seems low, underweight individuals

Decision Table

See the tradeoffs side by side

Criterion Mifflin-St Jeor (1990) Harris-Benedict Revised (1984)
ADA recommendation Recommended as primary formula No longer preferred
Accuracy (within 10% of measured) 82% of subjects 69% of subjects
Typical bias Slight underestimate (~2%) Overestimates by ~5%
Year published 1990 1984 (revised from 1919)
Uses body fat %? No No
Best alternative if body fat known Katch-McArdle (uses lean mass) Katch-McArdle (uses lean mass)

Verdict

Use Mifflin-St Jeor as your primary formula — it's what the ADA recommends and what our TDEE Calculator uses as its default. But run both and check the Formula Comparison tab to see the range. If they agree within 100 calories, you have a solid estimate. If they disagree by 200+, your profile is harder to estimate and you should calibrate with 2-3 weeks of real-world tracking. And if you know your body fat percentage, add it to enable Katch-McArdle — it bypasses the height/weight limitations of both formulas entirely.

FAQ

Questions people ask next

The short answers readers usually want after the first pass.

Mifflin-St Jeor vs Harris-Benedict: which is more accurate?
Mifflin-St Jeor. In the American Dietetic Association's 2005 comparative review (Frankenfield et al.), it predicted resting metabolic rate within 10% of measured values for 82% of subjects, versus 69% for Harris-Benedict. Harris-Benedict's error grows with body fat because it tends to overestimate by roughly 5%, so the gap widens for heavier individuals. The one case where Harris-Benedict may edge ahead is underweight people (BMI under 18.5). For everyone else, default to Mifflin-St Jeor and only reach for Katch-McArdle if you have a measured body fat percentage.
Can I just average the two formulas?
Yes, and that's reasonable. Our calculator shows the average across all formulas it runs. Averaging tends to reduce the error from any single formula's bias. That said, if you want one number, Mifflin-St Jeor is statistically more likely to be closer to reality.
What about the Katch-McArdle formula?
Katch-McArdle uses lean body mass instead of total weight, making it potentially more accurate for people who carry significantly more or less muscle than average. The downside: you need a body fat percentage measurement, which itself has measurement error. Our calculator includes it automatically when you provide body fat %.
Why do some apps use Harris-Benedict by default?
Legacy. Harris-Benedict has been around since 1919 and was the default in nutrition software for decades. Many apps and online calculators simply never updated. If your app gives you a BMR and it seems 100-200 calories higher than our calculator, it's likely using Harris-Benedict.

Sources & References

Related Content

Keep the topic connected

General fitness estimates — not medical advice. Consult a healthcare professional for medical decisions.