BMI vs Body Fat: Which Number Should You Use?
BMI is your weight in kilograms divided by your height in metres squared. Body fat percentage is the share of that weight made of fat. One takes two seconds, the other takes a tape measure, and they answer two different questions. BMI asks how heavy you are for your height. Body fat asks what the weight is made of.
Run both.
Two men, same BMI, 15.8 kg apart
Both stand 180 cm and weigh 88 kg. BMI of 27.2 each, overweight category each, identical to the decimal place.
The first trains four times a week and sits at 12% body fat, which works out to 10.6 kg of fat riding on 77.4 kg of lean mass. He is in excellent shape. The category is flatly wrong about him.
The second last trained fifteen years ago and carries 30%. That is 26.4 kg of fat on 61.6 kg of lean. Same BMI, 15.8 kg more fat, a completely different risk profile.
BMI reads them as twins. That is the real case against it.
Where BMI is right more often than people admit
The muscular exception is real, and it is also rare. If you have never trained seriously with weights, the odds are heavily against you carrying enough muscle to break the formula. Across the general adult population BMI tracks body fat at a correlation of roughly 0.7, which is loose for one person and perfectly serviceable for screening a population.
BMI is also free and close to idiot-proof. Put a tape measure an inch out of place and you walk away with a confidently wrong body fat figure, whereas a bathroom scale and a doorframe will hand you the same BMI every morning for the rest of your life. Two inputs, both unambiguous.
The BMI calculator gives you the score and, more usefully, the healthy weight range for your height. That range is often the better half of the output. What BMI measures walks through the four situations where it gives the wrong answer.
Where the percentage earns its two minutes
Any time the composition is the point. Training, dieting, past sixty, or parked within a kilo of a category line: in all four cases the percentage tells you what the scale keeps to itself. It is also the only one of the two that shows whether a diet is taking fat or muscle, which is the thing that decides whether the diet worked.
The body fat calculator uses tape circumference and lands within about 3 percentage points of a DEXA scan for most people. Close enough to track over months. Too loose to argue over a single point.
Normal weight obesity, the failure mode nobody mentions
A BMI of 23 sitting alongside 33% body fat reads healthy on every chart and carries the metabolic risk of someone considerably heavier. It turns up most in inactive adults past forty who have held the same weight for twenty years while quietly shedding muscle, because adults lose 3 to 8% of their muscle mass per decade after thirty.
Treat a healthy BMI in an untrained person as one input. This is exactly the case where two minutes with a tape measure pays for itself.
The third number, and the case for checking it first
Where the fat sits drives risk, and neither BMI nor body fat percentage can tell you where it is. Fat packed around your organs behaves very differently from fat under the skin of your hips.
Waist divided by height, target under 0.5, takes one measurement and predicts metabolic risk better than BMI does in most studies. Waist to height ratio covers the evidence, and visceral fat and why it matters more explains what the tape is picking up.
What to do with your five minutes
Check all three, once. BMI hands you the healthy weight range, body fat tells you what the weight is made of, and waist to height tells you where it has settled. Agreement means you have your answer. But if BMI says overweight while the other two look fine, go with the other two, and if BMI says healthy while the other two look poor, go with the other two again.
BMI is the one that gets overruled. Which is a fair summary of what it is for.
Scope and correction note, 29 October 2026: the ranges quoted here come from the American Council on Exercise and from published research, and they describe averages, not any individual. Nothing on this page is medical advice. If a figure here is wrong, tell us and we will correct it and date the correction. See the editorial policy and medical disclaimer.