BMI is not a measure of body fat. It is a ratio of weight to the square of height, and every useful thing it does — and every way it misleads — follows from that fact. It is a good screening instrument and a poor diagnostic one, and the distinction is not pedantry: the same number can mean opposite things for two people standing side by side.
The formula, and why it is squared
BMI = weight (kg) ÷ height (m)², giving units of kg/m². The BMI calculator applies exactly this and handles the unit conversions so you do not have to.
The square is not arbitrary. Body weight scales roughly with height linearly in adults — a person 10% taller weighs about 10% more — so dividing weight by height alone would leave a number that drifts upward with stature and systematically misclassifies tall people as heavier. Dividing by height² removes that trend: it is a statistical correction chosen because it makes adult BMI roughly independent of height, not because it arises from any physiological argument.
That origin explains both its usefulness and its limits. BMI was constructed as a population-level index by Adolphe Quetelet in the 1830s, decades before anything was known about body composition or fat distribution. It works as a rough population correlate of risk because heavier people do have higher mortality on average — but it works by averaging over the fact that the body can be fat or muscular at the same weight.
Where the cutoffs come from
The adult bands are widely reproduced and, contrary to a common belief, largely identical across major bodies:
- Underweight: below 18.5
- Healthy weight: 18.5 to 24.9
- Overweight: 25 to 29.9
- Obesity: 30 and above, with subclasses at 35 and 40
The WHO and the CDC both use 25 and 30 as the adult overweight and obesity thresholds. The CDC states plainly that BMI "is a screening measure" and "should be considered with other factors when assessing an individual's health" — that is the correct way to read the number, and it is the framing most popular coverage drops.
Worked example. A 1.80 m person weighing 90 kg has BMI = 90 ÷ (1.80)² = 90 ÷ 3.24 = 27.78. That falls in the 25–29.9 band, so it is classified as overweight under both the WHO and CDC criteria — the same conclusion either way, since the two agree on these numbers. The subclasses at 35 and 40 also matter for treatment decisions: a clinician looking at a BMI of 37 will react differently than at 34, which is information the four-category view hides.
Population differences, and why Asia uses lower cutoffs
BMI is a European-derived index applied to bodies that are not average European bodies, and the mismatch is measurable. At the same BMI, Asian populations typically carry a higher body fat percentage and a greater proportion of visceral fat than European populations, and their risk of type 2 diabetes and cardiovascular disease starts rising at lower BMI values.
A WHO expert consultation on appropriate BMI for Asian populations concluded that the standard cutoffs do not adequately reflect risk in many Asian groups and recommended additional, lower trigger points: 23 as the threshold for increased risk and 27.5 for high risk. Several countries and regions have since adopted versions of these. China uses 24 and 28 as its overweight and obesity boundaries, and the Western Pacific regional office applies roughly 23 and 25.
This matters for the example above. At a BMI of 27.78, that person is "overweight" under European cutoffs and already at or past the higher-risk threshold under Asian-specific ones. The same measurement, materially different implication. The percentage calculator is useful for expressing how far a value sits from a threshold rather than simply above or below it.
Where BMI fails outright
Muscular people. Muscle is roughly 1.06 g/cm³ while adipose tissue is around 0.94 g/cm³, so muscle weighs more per litre of volume. A powerlifter and a sedentary person of identical height and weight can carry 12% and 32% body fat respectively at the same BMI. The athlete is flagged as overweight while carrying very little fat; the sedentary person passes screening while carrying substantially more. This is the clearest demonstration that BMI is not measuring what people assume it measures, and it is why a body fat estimate is the better tool whenever one is available.
Short people. Squaring a smaller height produces a larger number, so for people well below average height the same weight yields a higher BMI than the equivalent frame at average height. Very tall people show the mirror-image effect, weaker but present.
Children. Body composition changes with growth and puberty, so a fixed adult table is meaningless. For ages 2 to 19, BMI is plotted against age- and sex-specific reference curves and read as a percentile — roughly the 85th percentile and above is the conventional threshold. A single number on an adult table is the wrong instrument for a child. BMI tables also do not apply during pregnancy.
What fills the gap: waist and waist-to-height ratio
What BMI cannot capture is where the fat sits, and that matters because fat around the abdomen and internal organs behaves differently from fat under the skin of the limbs. Waist circumference is the standard second measurement: the CDC cites higher risk at waist measurements above 35 inches (89 cm) for women and 40 inches (102 cm) for men.
Waist-to-height ratio is more useful still, because it needs only one extra measurement and removes the need for sex-specific cutoffs. The widely used threshold is WHtR ≥ 0.5 — keep your waist below half your height.
Worked example. A person 1.75 m tall with a 0.95 m waist has WHtR = 0.95 ÷ 1.75 = 0.543. Since 0.543 > 0.5, that crosses the usual threshold for central adiposity, indicating fat carried around the abdomen rather than the hips and thighs.
Compare two people at the same height and weight — say 1.75 m and 70 kg, both at BMI 70 ÷ 3.0625 = 22.86, comfortably "healthy" by BMI. A lifter in that range might be at 12% body fat with an 0.82 m waist, giving WHtR = 0.469, below the threshold. A sedentary person at 32% body fat might carry a 1.00 m waist, giving WHtR = 0.571, above it. Identical BMI, opposite readings from the ratio. That divergence is the whole argument for measuring more than one thing, and the body fat and metabolism guide goes deeper on the composition side.
When body fat is the better answer
For anyone whose BMI is high, the useful question is not "is this too heavy" but "how much of this is fat". For people classified as overweight or obese, body fat percentage is generally more informative than BMI, because it is the thing being tracked clinically. For people with a normal BMI, body fat percentage can still reveal the condition sometimes called skinny fat — normal weight, high fat, low muscle — which BMI is structurally incapable of detecting. The calorie calculator is often built on body composition for exactly this reason, and the trade-offs between estimation methods are covered in the next guide.
How to use the number honestly
Treat BMI as one input among several, and be explicit about its role. Compute it, note where the value falls, and then ask what it cannot tell you — because those are different questions. A BMI of 27.8 tells you this person weighs more relative to height than most of a reference population. It does not tell you their body fat, their fat distribution, their muscle mass, their blood pressure, or their blood markers. Any claim that treats the number as a diagnosis has skipped the part where the diagnosis happens.
Cutoffs also move. The values above reflect guidance current at the time of writing; clinical practice and national standards are revised periodically, so the latest published guideline should always be the authority.
This article is educational and is not medical advice. BMI, waist and body fat figures are screening references with wide individual variation and imperfect accuracy. Do not use any figure here to diagnose, treat or modify your health. Individual risk varies with age, sex, ethnicity, medical history, medication and body composition. Talk to a qualified clinician or registered dietitian before acting on any number, and seek care for symptoms rather than interpreting them with an index.
Frequently asked questions
What is a good BMI for adults?
For adults the standard cutoffs run underweight below 18.5, a healthy weight from 18.5 to 25, overweight from 25 to 30, and obesity at 30 or above. These are screening bands rather than diagnoses, and several national bodies use different figures, so the numbers should be read alongside waist measurement and other health markers.
Why can two people with the same BMI have very different body fat?
Because BMI cannot tell muscle from fat. Muscle is denser than fat, so at an identical weight and height a person with more muscle will register a higher BMI at a lower body fat percentage. This is why BMI misclassifies athletes as overweight and can miss an elevated fat percentage in a sedentary person with a normal BMI.
Why are lower BMI cutoffs used for Asian populations?
Because body fat percentage at a given BMI is systematically higher in Asian populations, so the risk of type 2 diabetes and cardiovascular disease rises at lower BMI values than in European populations. WHO expert consultation therefore recommended additional lower trigger points, and several countries now use cutoffs around 23 and 27.5.
Is BMI useful for children?
Not on its own. Children and adolescents are growing, and body composition changes with age and puberty, so a single adult cutoff is misleading. For ages 2 to 19, BMI is interpreted against age- and sex-specific reference percentiles rather than fixed bands, which is why a paediatric percentile chart is used instead of an adult table.