A TDEE calculator with body fat percentage uses the Katch-McArdle formula instead of (or alongside) Mifflin-St Jeor. The advantage: lean body mass is a much better predictor of metabolic rate than total body weight is — because muscle and organs burn calories, while fat is metabolically near-inert. The disadvantage: you need an accurate body fat number, and most people don't have one. This page walks through when the body fat input actually improves accuracy, how to estimate it well enough to use, and when to ignore it entirely.
TDEE Calculator
Total Daily Energy Expenditure — your real maintenance calories.
Biological sex
BMR formulas use biological sex, not gender identity.
Leave blank if unsure — we'll fall back to Mifflin-St Jeor (the standard).
Most people overestimate this. When in doubt, drop one tier.
The Mifflin-St Jeor equation predicts your BMR from age, sex, height, and weight. It works because for the average adult, body composition is reasonably predictable from those inputs — a 30-year-old 170-pound man at 5'10' has a fairly typical ratio of muscle to fat.
Katch-McArdle skips the proxies and uses lean body mass directly. Its core: BMR = 370 + (21.6 × lean mass in kg). Body fat is metabolically near-inert (about 3 kcal/lb/day, almost negligible), so once you know your lean mass, you've isolated the part of your body that's actually doing the burning.
For the average person at typical body composition, both formulas land within 50–100 kcal of each other. The difference matters in two scenarios: people who are unusually muscular (bodybuilders, athletes) and people who are unusually fat (BMI 35+). In both cases, Mifflin's assumptions break down, and Katch-McArdle delivers a meaningfully more accurate number.
When the two formulas disagree (and why)
Same person profiles, different body compositions. See where Mifflin and Katch-McArdle diverge.
Profile
Weight
Body fat %
Mifflin BMR
Katch-McArdle BMR
Δ
Average male, 30, 5'10"
175 lb
20%
1,710
1,743
+33
Muscular male, 30, 5'10"
175 lb
10%
1,710
1,914
+204
Heavyset male, 30, 5'10"
230 lb
35%
2,061
1,832
−229
Average female, 30, 5'5"
140 lb
25%
1,371
1,395
+24
Athletic female, 30, 5'5"
140 lb
15%
1,371
1,535
+164
Heavyset female, 30, 5'5"
200 lb
40%
1,640
1,562
−78
Notice: Mifflin under-estimates BMR for muscular bodies and over-estimates BMR for high body fat bodies. Katch-McArdle corrects both directions.
When to use body fat input vs ignore it
Practical guidance:
USE body fat input if you've had a DXA scan, BodPod, hydrostatic weighing, or a high-quality multi-frequency BIA in the last 6 months. Katch-McArdle will give a meaningfully better estimate.
USE body fat input if you're visibly athletic (low body fat, visible muscle definition) — Mifflin will under-estimate your BMR.
USE body fat input if you're carrying significant body fat (BMI 32+) — Mifflin will over-estimate.
IGNORE body fat input if your only number comes from a bathroom scale BIA. These devices are notoriously inaccurate (±5–10% body fat is normal error), and a wrong body fat input produces a worse estimate than just letting Mifflin do its thing.
IGNORE body fat input if you don't know your body fat at all. Default to Mifflin. Don't guess — guessed body fat is worse than no body fat input.
How to estimate body fat without lab equipment
If you don't have access to a DXA but want to use Katch-McArdle, you can estimate body fat from visual cues plus a tape measurement. None of these are perfect, but they're better than a bathroom scale's BIA reading.
Visual estimation. Compare yourself to body fat reference charts (search 'body fat percentage visual chart'). Look for the photo that most closely matches your current physique from the front and side. Accuracy: ±3–5% for honest estimators, ±5–10% for self-flattering ones.
US Navy method. Requires tape measurements: neck circumference, waist circumference (men), and additionally hips (women). Plug into the Navy formula (available in many calculators). Accuracy: ±3% on average, with bigger errors at the extremes.
3-site or 7-site skinfold caliper. If you have a caliper and a willing partner. Accuracy: ±2–3% if done correctly.
If you're between estimates, round up your body fat by 1–2% (the typical self-estimation bias is toward leaner numbers). Then plug the rounded estimate into the calculator and use that as your starting point. Re-verify with the eat-flat experiment described elsewhere on this site.
Field note from Sukie
The friend who got radically different numbers from three sources
A friend (R) wanted to know his body fat. He tested three ways in one week. The bathroom scale BIA said 14%. The hand-held BIA at his gym said 18%. The DXA scan at a sports medicine clinic said 22%. Three numbers, eight-percentage-point spread, same body in the same week.
The DXA is the most accurate of the three. The bathroom scale was off by 8 percentage points — not unusual for those devices, which mostly measure water content and infer body fat. He'd been using the bathroom scale's 14% number for years and assuming Katch-McArdle was giving him a better estimate than Mifflin. The 14% input was actually giving him a worse estimate than just using Mifflin without body fat at all.
Moral: if you're going to use the body fat input in this calculator, use a number from DXA, BodPod, hydrostatic, or a well-executed caliper test. Anything less and you're feeding the formula noise.
— Sukie
What the calculator does internally
When you enter body fat in the calculator above, here's what happens:
If body fat is provided: it computes lean body mass (weight × (1 − body fat %)) and runs Katch-McArdle. The output is labeled 'Katch-McArdle BMR.'
If body fat is not provided: it runs Mifflin-St Jeor. The output is labeled 'Mifflin BMR.'
If both are computed (when you provide body fat): we show both numbers and use Katch-McArdle as the canonical BMR. The Mifflin number is shown for comparison so you can see how the two differ for your specific body.
From the BMR, we multiply by your activity factor to get TDEE, and apply your goal modifier (cut/maintain/bulk) to get target calories. Macros are computed from the target calories using fixed protein-per-kg, fixed fat percentage, and remaining calories to carbs.
Edge cases where Katch-McArdle clearly wins
Three populations where the body-fat-aware formula is genuinely better:
1. Competitive bodybuilders during contest prep. Body fat drops to 5–8% (men) or 12–15% (women), and lean body mass is the dominant determinant of metabolism. Mifflin systematically under-predicts BMR for these athletes by 200–400 kcal. Katch-McArdle gets within 50 kcal. Coaches in this space use Katch-McArdle almost exclusively.
2. Powerlifters and strongman competitors at high bodyweights. A 280 lb powerlifter at 22% body fat has way more lean mass than a typical 280 lb person, and Mifflin treats them the same. Katch-McArdle correctly accounts for the muscle.
3. Bariatric surgery patients during weight loss. Lean body mass is preserved at a different rate than fat mass, and Mifflin's age/weight/height inputs lose their predictive power as body composition shifts dramatically. Specialists working with these patients use Katch-McArdle plus serial DXA scans.
For everyone else — gym-going adults at typical body composition — Mifflin is fine.
How body fat measurement error propagates into TDEE error
Katch-McArdle is more accurate than Mifflin when body fat is accurate. The catch: most people don't have an accurate body fat number, and the formula doesn't know that. A wrong body fat input produces a wrong BMR, which produces a wrong TDEE.
The propagation isn't terrible but isn't trivial either. Each percentage point of body fat error shifts the BMR estimate by roughly 10–20 kcal for a typical adult. A 5-point error (which is normal for bathroom-scale BIA and self-estimated visual checks) shifts BMR by 50–100 kcal. Multiplied through the activity factor, that's a 75–175 kcal error in TDEE — about the same as a typical Mifflin error.
So the practical rule: only use Katch-McArdle if you have a body fat number accurate to within 3%. DXA, BodPod, or a careful 7-site caliper test by a trained tester all qualify. A bathroom scale's body fat reading does not — it's so noisy that running Katch-McArdle on it is worse than running Mifflin without it.
If you're between methods, Mifflin is a stronger default than Katch-with-bad-body-fat. Switch to Katch only when you've leveled up your body fat measurement, not before.
Frequently asked questions
Is Katch-McArdle always more accurate than Mifflin-St Jeor?+
Only if your body fat input is accurate. With a guessed or scale-BIA body fat number, Katch-McArdle can be less accurate than Mifflin. Use Katch-McArdle when you have DXA, BodPod, or quality caliper data — otherwise stick with Mifflin.
What's the most accurate way to measure body fat?+
DXA scan is the practical gold standard for body composition — costs $50–150 in most US markets. BodPod and hydrostatic weighing are also lab-grade. Skinfold calipers performed by a trained tester are next best. Bathroom scale BIA is the least accurate consumer option.
Can I use Katch-McArdle if my body fat estimate is just a guess?+
No — a wrong body fat number produces a worse BMR estimate than no body fat input at all. If you can't measure body fat accurately, use Mifflin without the body fat field.
How often should I update my body fat for the calculator?+
Every 3–6 months during active body composition phases (cutting or bulking). Less often during maintenance. Body fat changes faster than you'd think when you're in a meaningful deficit or surplus.
Does the calculator use the latest body fat input automatically?+
It uses whatever number you enter at the time of calculation. Re-enter your current body fat each time you recompute, especially after meaningful weight changes.
Why does my BIA scale give a different body fat reading day to day?+
BIA measures water content and infers body fat. Hydration, time of day, food intake, and even ambient temperature affect the reading. Day-to-day variation of 1–3% is normal and doesn't reflect real body composition change.