What BMI Does and Doesn't Tell You

By Jesse · Last updated August 20, 2026

Body Mass Index is the most widely used weight metric in the world and one of the most over-interpreted. It is genuinely useful for the job it was built for, and actively misleading the moment you ask it to do more. The distinction is worth getting right, because most arguments about BMI are really arguments about which question is being asked.

What BMI actually is

Weight in kilograms divided by height in metres squared. That is the whole formula, and its inputs are the whole story: BMI knows your height and your weight, and nothing else about you. Find yours with the BMI calculator, or read the value off the BMI chart by height and weight.

The ratio dates to the 1830s, when the Belgian statistician Adolphe Quetelet observed that weight in a population scales roughly with the square of height. It was never meant to assess an individual — Quetelet was describing populations. The name "body mass index" came much later, from Ancel Keys and colleagues in 1972, who compared the available height-weight indices and concluded this one was the best available proxy for population studies. Both origins point the same way: it is a screening statistic, and it has been doing a job it was not designed for ever since.

What BMI does well

Three real strengths, none of them trivial. It is cheap — two measurements anyone can take, no equipment, no technique to get wrong, which is why it scales to national surveys. It is comparable — the same number means the same thing across studies and decades, which is what makes long-run public health data possible at all. And across large groups it tracks risk: higher average BMI is associated with higher rates of type 2 diabetes, cardiovascular disease, and several other conditions.

Notice that all three are statements about groups. That is not a weakness of BMI; it is its design specification.

Where it misleads

The core limitation follows directly from the inputs: BMI measures mass, not composition, and it cannot see shape at all. It cannot distinguish muscle from fat, or fat on the abdomen from fat on the hips.

That is measurable rather than rhetorical. Hold height and weight fixed — so BMI is pinned at 24.9 (Normal weight) — and vary only the tape measurements. The body-fat estimate moves across 18.9 percentage points and 4 separate categories while BMI reports the same number throughout. Run it the other way and the same thing happens in reverse: at fixed tape measurements, body weight can move 30.6 kg across 3 BMI bands without the body-fat estimate changing at all.

Four consequences, all of which show up in practice:

• Muscular people read as overweight. A lean, heavily trained athlete can sit in the overweight or obese range on BMI despite low body fat, because muscle is dense and BMI only sees the total.
• A normal BMI can hide high body fat.Low muscle mass with more fat can produce a textbook-normal BMI alongside a body-fat percentage that is not — the case sometimes called being "skinny fat."
• Fat distribution is invisible to it. Abdominal fat carries more metabolic risk than fat on the hips and thighs, and BMI cannot tell the difference, because circumference is not one of its inputs.
• The same number means different things across groups. Age, sex, and ethnicity all shift the relationship between BMI and body fat, so fixed cut-offs are approximations rather than universal boundaries.

Four bodies where BMI and body fat disagree

These are plausible bodies, chosen to make the point rather than sampled from a population. Each row is one person, classified twice — by BMI from height and weight, and by the U.S. Navy circumference method from height and tape:

Read the first two rows together. The lighter man is called normal weight by BMI and 26% body fat by the tape. The man 23 kg heavier, at the same height, is called obese by BMI and 14.5% by the tape. The heavier man is estimated to carry 11.5 percentage points less fat. The two methods are not disagreeing about degree; they have swapped places.

The disagreement is structural. BMI reads height and weight. The Navy formula reads height, neck, waist and hips — and never reads weight at all. The only input they share is height, which is why they can be made to contradict each other almost at will. We worked through what that structural gap does, including which method to believe in which situation.

The cut-offs are conventions, not biology

The World Health Organization sorts adults into underweight (below 18.5), normal (18.5 to 24.9), overweight (25 to 29.9), and obese (30 or above). Those are round numbers chosen where population risk curves start bending, not physiological switches — nothing changes about your body between a BMI of 24.9 and 25.0.

The clearest evidence that they are conventions is that they have been revised for different populations. A WHO expert consultation examined evidence that people of Asian descent tend to carry a higher percentage of body fat and greater cardiovascular risk at a given BMI, and discussed lower action points — around 23 — for those groups. If a threshold can be moved by 2 points on the basis of population data, it is a policy instrument, not a law of nature.

What to look at alongside it

BMI is one cheap dimension of a bigger picture. Three companions that are nearly as cheap:

• Waist-to-height ratio. Measure your waist and divide by your height, in the same units. Keeping it below 0.5 — waist under half your height — is a widely recommended target, and it captures the abdominal fat BMI is blind to. It needs one extra measurement and no equipment beyond a tape.
• A body-fat estimate. Even a rough one from the body fat calculator tells you something BMI structurally cannot. Treat the absolute percentage with caution and the direction of change as the useful signal.
• Your own trend. A single reading against a fixed threshold is the least informative way to use any of these numbers. The same measurement taken monthly under the same conditions is worth more than a more sophisticated one taken once.

How to use it properly

Treat BMI as a first screen with one job: telling you whether a closer look is warranted. A BMI outside the normal range is a reason to measure your waist, estimate body composition, and look at blood pressure and blood markers — not a diagnosis, and not a verdict on your health. A BMI inside the normal range is reassuring but not exonerating, for exactly the same reason. Check yours with the BMI calculator, then put it in context rather than treating it as the conclusion.

Common questions

This is general information for healthy adults, not medical advice. For a proper assessment of your weight and health, talk to a doctor. See our methodology for how we use these formulas.