TDEE Calculator with Body Fat %

Compare Mifflin-St Jeor and Katch-McArdle side by side. If you know your body fat, Katch is the more accurate of the two — especially at the extremes of body composition.

Your Stats

Gender

Optional. If provided, Katch-McArdle results appear alongside Mifflin-St Jeor.

Your Results

Mifflin-St Jeor
—
Enter your stats to compare both models.
Katch-McArdle
—
Add your body fat % above to see this.

Why a body-fat-aware TDEE is more accurate

BMR — the calories you burn just existing — is mostly driven by lean tissue. Muscle, organs, brain, and bone are metabolically active; body fat is not, at least not nearly to the same degree. Two people who weigh exactly 180 lb can have very different BMRs depending on how much of that 180 lb is lean and how much is fat.

The Mifflin-St Jeor equation has no way to know that. It treats weight as weight. The Katch-McArdle equation strips out the body fat and uses lean body mass directly, which is why it's the standard for athletes, bodybuilders, and anyone tracking serious body recomposition.

The Mifflin-St Jeor equation

Mifflin-St Jeor estimates BMR from weight, height, age, and gender:

  • Men: BMR = 10 × kg + 6.25 × cm − 5 × age + 5
  • Women: BMR = 10 × kg + 6.25 × cm − 5 × age − 161

For someone with average body composition this is within ~10% of measured BMR — good enough for ~80% of users and the reason it's the default on the generic TDEE calculator.

The Katch-McArdle equation

BMR = 370 + 21.6 × lean body mass (kg)

That's it. One number — lean mass — does all the work. To get lean body mass, multiply your weight by (1 − body fat / 100). A 180-lb person at 20% body fat has 144 lb of lean mass, which converts to ~65.3 kg.

When each model wins

  • Use Mifflin when: you don't have a reliable body-fat number. Mifflin's average-person assumption is fine for average bodies.
  • Use Katch when: you have a measured body-fat estimate (DEXA, hydrostatic, even a calibrated bioimpedance scale), you're notably lean or notably high-fat, or you carry more muscle than average.
  • Both lose when: the input is wrong. A guessed body fat fed into Katch isn't more accurate than Mifflin — it's just more sensitive to the guess.

How to estimate your body fat

DEXA is the gold standard but costs $50–$150 a session. Hydrostatic weighing and Bod Pod are also reliable but rarely accessible. Smart scales using bioimpedance are typically within ±4% — fine for tracking change over time, less useful for a single absolute reading. Visual estimation against photo charts is a free and surprisingly decent fallback for anyone outside the extremes.

Frequently asked questions

Which formula is more accurate — Mifflin or Katch-McArdle?

It depends on what you know about yourself. Mifflin-St Jeor uses only weight, height, age, and gender — it's accurate within ~10% for the average person. Katch-McArdle uses lean body mass (your weight minus body fat) and is more accurate when you actually know your body fat percentage. The leaner you are or the more muscular, the more Katch wins.

Do I need a DEXA scan to use Katch-McArdle?

No, but the closer your body-fat number is to reality, the closer Katch will be. DEXA and hydrostatic weighing are gold standard. Bioimpedance scales are roughly ±4%. Visual estimation from photos is fine for ballpark — most people can pin themselves within a 5% range.

Why do the two numbers differ so much when I'm very lean or very heavy?

Mifflin doesn't know your body composition. A 200-lb lean lifter and a 200-lb sedentary person look identical to it. Katch sees the lean lifter's higher lean mass and gives a higher BMR — usually correctly. The reverse is true for someone with high body fat: Mifflin overestimates because it doesn't discount the metabolically less active fat mass.

Which TDEE should I actually use for my cut?

If you know your body fat to within ~3%, use Katch. If you're guessing your body fat, use Mifflin — a guess wrapped in a more sensitive formula isn't more accurate. Either way, real-world feedback (the scale and the mirror over 2–3 weeks) beats either model.

Related tools