Your basal metabolic rate is the foundation of every macro number you will set for the rest of your training career. If the foundation is wrong, everything built on top of it is wrong too.

BMR is the number of calories your body burns at complete rest just to keep you alive — breathing, circulating blood, maintaining body temperature, powering your brain. Every macro calculator then multiplies this number by an activity factor to estimate your total daily energy expenditure (TDEE) and splits the result into protein, carbs, and fat targets. If the BMR estimate is off by even 100 calories, the error compounds across every day type, every week, every month.

There are three widely used BMR formulas in the nutrition world: Mifflin-St Jeor, Harris-Benedict, and Katch-McArdle. They give different results for the same person, and the differences matter most for athletes. This article compares them head-to-head, explains why Katch-McArdle is the superior choice for athletic populations, and shows you exactly what the difference means in practice.

If you have not already read our guide on how to set macros for CrossFit athletes, start there for the full system. Here, we are zooming in on the single most important input to that system.

The Three BMR Formulas at a Glance

All BMR formulas are regression equations derived from indirect calorimetry studies — meaning researchers measured people's actual oxygen consumption and carbon dioxide production, then fitted an equation to predict those measurements from easy-to-collect variables like weight, height, age, and sex.

Mifflin-St Jeor (1990)

Developed by Mifflin and St Jeor at the University of Nevada and published in the American Journal of Clinical Nutrition in 1990 [1], this formula is currently the most widely recommended equation for the general population. It uses total body weight, height, age, and a sex-based constant:

Male: BMR = 10w + 6.25h − 5a + 5
Female: BMR = 10w + 6.25h − 5a − 161

Where w = weight in kg, h = height in cm, a = age in years

Accuracy claim: Predicts BMR within ±10% of measured values in about 80% of the general population [2]. Its main strength is that it does not require body composition data, making it practical for clinical and mass-market use.

Harris-Benedict (1919, revised 1984)

The original Harris-Benedict equation was published in 1919 and revised by Roza and Shizgal in 1984 [3]. It is the oldest of the three and the least accurate for modern populations, in part because the original study subjects were mainly young, lean Caucasian men from a century ago:

Male: BMR = 88.362 + (13.397w) + (4.799h) − (5.677a)
Female: BMR = 447.593 + (9.247w) + (3.098h) − (4.330a)

Where w = weight in kg, h = height in cm, a = age in years

Multiple validation studies have found that Harris-Benedict overestimates BMR by 5-15% in modern populations, making it the least desirable option for anyone serious about precision [2][4].

Katch-McArdle (1993)

Published by Frank Katch and William McArdle in their textbook Introduction to Nutrition, Exercise, and Health [5], this formula takes a fundamentally different approach. Instead of total weight, height, and age, it uses lean body mass alone:

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

The formula is identical for males and females. It does not need height or age because lean body mass already captures the metabolically active tissue that drives resting energy expenditure. Two people with the same lean body mass will have the same Katch-McArdle BMR regardless of their height, age, or sex — and research supports that this is more accurate than the alternative [6].

Why Katch-McArdle Wins for Athletes

The fundamental insight of the Katch-McArdle formula is that muscle tissue is metabolically active and fat tissue is not. A kilogram of muscle burns roughly 13 kcal/kg/day at rest, while fat tissue burns only about 4.5 kcal/kg/day [7]. Total weight cannot distinguish between these two tissue types, but BMR depends almost entirely on how much muscle you carry.

For the general population, this distinction matters less because the ratio of muscle to fat is relatively consistent across people of the same weight, height, and age. But athletes deliberately manipulate this ratio through training. A CrossFit athlete may have 12% body fat while a sedentary person of the same weight may have 22% body fat. Their total weight is identical, their Mifflin-St Jeor BMR is identical, but their actual metabolic rates differ substantially.

The Same Weight, Different BMR

Here is the key comparison. Take two athletes who both weigh 175 lbs (79.4 kg) and are both 30 years old and 5'10" tall:

Mifflin-St Jeor gives both athletes the same BMR because it only knows their weight, height, and age. But Athlete A carries 17.5 more pounds of metabolically active muscle. Katch-McArdle captures this: a 173 kcal/day difference between two people with the same scale weight.

Over a year, that is a discrepancy of 63,145 calories — roughly 18 pounds of body mass. If Athlete A followed a Mifflin-St Jeor-based plan designed for Athlete B, they would be chronically underfed by ~83 kcal/day despite being at a higher calorie total. Their performance would suffer, and they would struggle to gain or maintain muscle mass despite following the numbers.

Why the Gap Exists

The Mifflin-St Jeor formula was validated on a sample that was 64% female and 80% Caucasian, with a mean BMI of 28.8 [1]. The study deliberately recruited a general population sample to develop a formula applicable to the broadest possible range of people. It succeeds at that goal but necessarily flattens the body composition variation that matters most for athletes.

Multiple validation studies have confirmed this limitation. Frankenfield et al. (2005) compared all three formulas against measured resting metabolic rate in 130 adults and found that Katch-McArdle was the most accurate for lean individuals, while Mifflin-St Jeor performed best for overweight and obese individuals who lacked body composition data [2]. The takeaway is clear: use the formula that matches your population.

Worked Comparison: Our Standard Test Case

Let us run through the full calculation for our standard test case: a 175 lb male at 15% body fat, 30 years old, 5'10" tall.

Mifflin-St Jeor

Note: the 1,760 figure above uses exact metric conversion. With the common rounded inputs of 79 kg and 178 cm, you often see ~1,790 quoted. The exact value is 1,760.

Katch-McArdle

The difference for this athlete: 68 kcal/day in favour of Katch-McArdle. Applied to a CrossFit training day (activity multiplier 1.725), that becomes 3,153 kcal vs 3,036 kcal — a 117 kcal/day TDEE gap. Over a week with three CrossFit days, two strength days, and two rest days, the cumulative difference is roughly 500 calories. Not dramatic for a single week, but over a 12-week training block it reaches 6,000+ calories, which is the difference between maintaining weight and losing or gaining a noticeable amount of body mass.

How the Gap Changes with Body Fat

The table above holds total weight constant at 175 lbs (79.4 kg) and varies only body fat percentage. The two formulas converge around 18-20% body fat. Below that, Mifflin-St Jeor under-estimates true BMR for lean athletes. Above that, it over-estimates. This is precisely what the Frankenfield validation study predicted [2].

For our target audience — CrossFit and hybrid athletes who typically fall in the 10-18% body fat range as males or 18-28% as females — the Katch-McArdle formula gives a systematically more accurate result because it accounts for the above-average muscle mass these athletes carry.

When Mifflin-St Jeor Is Acceptable

Mifflin-St Jeor is not a bad formula. It is the best available option for the general population precisely because it does not require body fat data. If you do not know your body fat percentage, Mifflin-St Jeor will give you a reasonable BMR estimate, and most of the error will fall within the ±10% range that is considered clinically acceptable.

Use Mifflin-St Jeor when:

• You do not have a reliable body fat measurement. Visual estimates and consumer-grade bioelectrical impedance scales are often inaccurate. Without a DEXA scan, caliper measurement by a qualified professional, or a reliable navy tape method reading, your body fat percentage may be off by 3-5 points — which defeats the advantage of Katch-McArdle.
• You are in the general population range (18-25% BF for men, 25-32% for women). In this range, the two formulas converge within 50-100 kcal, which is within the measurement error of any BMR estimation method.
• You are using BMR for clinical weight management rather than athletic performance. For general weight loss, a 100 kcal error is negligible compared to the variance introduced by daily activity, meal timing, and adherence.

But if you are a CrossFit athlete, a powerlifter, a runner, or any hybrid athlete who trains multiple times per week with varying intensity, the Katch-McArdle formula is the better choice. Your body composition is different from the general population, and your fuel requirements are too precise to leave to a formula that ignores your muscle mass.

How HYBRID.MACROS Uses Katch-McArdle

The HYBRID.MACROS calculator uses Katch-McArdle exclusively because every user of the tool is assumed to be an athlete or someone training with athletic intent. If you are here, you are not the general population. You are someone who tracks their training, cares about body composition, and wants nutrition precision that matches their effort in the gym.

Here is how the BMR flows through the full calculation:

1. Input: Your weight, body fat percentage (or estimated BF%), and unit preference (lbs/kg).
2. BMR: 370 + (21.6 × lean body mass in kg) — the Katch-McArdle formula.
3. TDEE per day type: BMR × activity multiplier (Rest 1.2, Strength 1.55, CrossFit 1.725, Double 1.9).
4. Goal adjustment: Performance maintain (TDEE), Cut (−500 kcal), Bulk (+300 kcal).
5. Macro split: Protein from LBM, carbs from total weight, fat as residual (with a hormonal health floor).

Because we base protein on lean body mass in pounds and carbs on total body weight in kilograms, your entire macro prescription is tied to your actual body composition, not a population average. For a full walkthrough of every step with examples, read the complete guide to setting CrossFit macros.

For more detail on protein specifically, including how lean mass-based protein targets compare to total-weight recommendations, see our guide How Much Protein Do CrossFit Athletes Need?

Does the Formula Choice Actually Matter in Practice?

Short answer: yes, but only if everything else in your nutrition is dialled in.

If you are inconsistent with your food tracking, eat out frequently, or do not measure portions, the 68-120 kcal/day difference between BMR formulas is noise compared to your overall tracking error. But if you are an athlete who weighs food, tracks every meal, and targets specific macronutrient numbers for performance or body composition goals, the formula matters.

Here is a rule of thumb: use the formula that matches your access to data.

• If you have a reliable body fat percentage, use Katch-McArdle. It will be more accurate for your population.
• If you do not have body fat data, use Mifflin-St Jeor (but recognize that your BMR estimate may be off by 50-150 kcal depending on your actual composition).
• Harris-Benedict should be retired for athletic use. It was designed for a different population in a different century.

For those who want maximum precision, get a DEXA scan or a 7-site caliper measurement from a qualified professional every 8-12 weeks. Update your body fat percentage in the HYBRID.MACROS calculator, and your BMR, TDEE, and macros will automatically adjust to reflect your changing body composition. That feedback loop — train, measure, recalculate, fuel, repeat — is what separates guesswork from evidence-based nutrition.