BMI Calculator: The Quetelet Formula and What It Does Not See
Body mass index is weight divided by height squared. This page runs that calculation, returns the standard WHO adult category, and explains what the number is and is not evidence of. The same arithmetic is used in population studies and on most health-screening forms, but the page also states the reasons a single BMI should not be read as a diagnosis.
Results appear here as you type.
Worked examples
These are produced by the same function your browser just ran, computed when this page was built.
| Case | Output |
|---|---|
| Man, 178 cm, 82 kg | BMI: 25.9; Category: Overweight; healthyRange: 18.5-24.9; boundaryWeightKg: 9.1; outOfRange: false |
| Woman, 165 cm, 68 kg | BMI: 25; Category: Normal range; healthyRange: 18.5-24.9; boundaryWeightKg: 8.2; outOfRange: false |
| Man, 70 in, 180 lb | BMI: 25.8; Category: Overweight; healthyRange: 18.5-24.9; boundaryWeightKg: 9; outOfRange: false |
| Woman, 64 in, 140 lb | BMI: 24; Category: Normal range; healthyRange: 18.5-24.9; boundaryWeightKg: 8; outOfRange: false |
How it is calculated
BMI = weight in kg / (height in m)^2 = 10000 x weight(kg) / height(cm)^2 when centimetres are entered. The same ratio in imperial units reduces to the identical number because the conversion constants cancel.
Take a man weighing 82 kg at 178 cm. His height in metres is 1.78, squared that is 3.1684, and 82 divided by 3.1684 gives 25.9. The shortcut with centimetres is 10000 x 82 / (178 x 178) = 25.9, which is the same answer with no decimal shift. The unit selector only changes what you type in; the number it produces is unitless and comparable across studies.
Formula source: Keys A et al., J Chronic Dis 1972;25(6-7):329-43 (PMID 4650929, doi 10.1016/0021-9681(72)90027-6). Adult BMI classification cut-offs from WHO Tech Rep Ser 2000;894 (PMID 10834449).
Reference data on this page
Sources: Keys A, Fidanza F, Karvonen MJ, Kimura N, Taylor HL. "Indices of relative weight and obesity", J Chronic Dis 1972;25(6-7):329-43, doi 10.1016/0021-9681(72)90027-6, PMID 4650929 - the paper compares relative-weight indices and shows that weight divided by height squared correlated well with body fatness while being largely independent of height, which is why this index was favoured over alternatives such as weight/height or weight/height cubed. World Health Organization. "Obesity: preventing and managing the global epidemic", WHO Tech Rep Ser 2000;894:1-253, PMID 10834449 - the report adopts BMI cut-offs for underweight, normal range, overweight and obesity classes in adults, and discusses the limitations of BMI as a screening tool.
https://pubmed.ncbi.nlm.nih.gov/4650929/ · retrieved 2026-10-03 · citation checked
Limits of this tool
BMI is a ratio, not a body-composition measurement. Two people at BMI 27 can differ by ten points of body fat if one is muscular and the other is not, because the formula knows only weight and height. The cut-points printed on this page are screening thresholds from population studies, not diagnoses of health or disease for any particular person. The WHO report is explicit that the same cut-points may not apply to all ethnic groups, and the 1972 index paper was comparing indices, not declaring BMI a measure of adiposity with a known per-person error. This page does not convert BMI into a risk percentage, because that conversion does not exist in the sources cited here.
What the number is actually measuring
BMI is weight divided by height squared. That is the whole arithmetic, and its usefulness comes from being the same number in every study that uses it. A population whose average BMI rises from 24 to 25 over ten years is a real signal, because the measurement protocol has not changed.
What it is not measuring is body composition. The denominator is height squared, not fat mass. A person with a large frame and a lot of lean tissue can land in the same BMI band as someone carrying excess fat, and the number cannot tell the difference. The 1972 comparison found that weight/height squared correlated well with body fatness across groups, which is why it was preferred over simpler indices; that is a statement about populations, not about individuals.
The consequence for this page is simple: the calculator prints the ratio and the WHO band, and it refuses to label that band a diagnosis. If the result is 31, the honest reading is "this person falls in the obesity class I range on the published screening scale", not "this person is unhealthy" or "this person has X percent body fat".
Where the adult cut-offs come from
The WHO 2000 technical report is the source most pages cite when they print underweight, normal, overweight and obesity classes. Its adult cut-points are 18.5, 25, 30 and 35, with a further split at 40 for obesity class III. Those values were selected from evidence linking BMI to mortality and to cardiometabolic disease in mainly white populations, and the report notes explicitly that the relationship between BMI and risk differs by ethnicity.
The report also distinguishes between the index as a screening tool and as a diagnostic threshold. Screening tools are meant to flag people who would benefit from further assessment, not to assign a disease label from a single number. That distinction matters because BMI bands are often reproduced on clinic walls without the qualifying sentence.
For children, the WHO growth standards use centile bands rather than adult cut-offs, and the input range on this page starts at age 16 and height 120 cm so that nobody treats the adult classification as a paediatric rule.
Why the same BMI reads differently on different bodies
The ratio removes most of the dependency on height, which was the point of the 1972 comparison, but it removes only the linear height effect. A taller person with the same BMI has more absolute fat mass and more lean mass, because both scale with height. Whether that difference matters depends on what the number is being used for.
Sex and age shift the interpretation without appearing in the formula. Women carry proportionally more fat mass than men at the same BMI, and the difference is large enough that the body-fat-from-BMI page on this site adds a separate sex term to its estimate. Older adults lose height through disc compression and can gain fat while their BMI stays flat or even falls, which is why BMI alone under-reads some changes in body composition after retirement age.
Athletic or resistance-trained populations are the other common mismatch. Muscle is denser than fat, so a person with a high lean-mass fraction can sit in the overweight or obesity bands on the BMI scale while carrying a body-fat percentage that would be classified as normal by skinfold or bioelectrical impedance. The BMI page cannot see muscle, and it cannot see visceral fat, which is the compartment most strongly linked to cardiometabolic risk.
What to use alongside this number
A waist circumference or waist-to-height ratio adds information about fat distribution, which BMI does not contain. Two people at BMI 27 can differ in metabolic risk if one stores fat viscerally and the other subcutaneously, and a tape measure is the cheapest way to separate those cases.
If the number matters to a medical decision, the next step is a measurement that actually samples the compartment in question: blood pressure, lipids, glycated haemoglobin, or a body-composition method with a known error. None of those is a better population screen than BMI for its original purpose, and all of them are more informative about the individual standing in front of you.
For tracking change over months in the same person, BMI has one genuine strength: it is cheap, repeatable and insensitive to daily fluid shifts. A trend that moves by more than about one unit under consistent conditions is real; smaller oscillations are measurement noise.
Questions
What is a healthy BMI?
The WHO adult screening ranges are 18.5-24.9 for the normal range, 25-29.9 for overweight, 30-34.9 for obesity class I, 35-39.9 for obesity class II, and 40 and above for obesity class III. These are population screening thresholds, not diagnoses of health for a particular person.
Why does BMI use height squared?
The 1972 Keys comparison found that weight divided by height squared correlated best with measured body fatness across a range of populations while being largely independent of height, which is why this index was preferred over weight/height or weight/height cubed.
Is BMI accurate for muscular people?
No, and that is a property of the index rather than a bug in this calculator. Muscle is denser than fat, so a muscular person can register as overweight or obese on the BMI scale while having a normal body-fat percentage. BMI does not distinguish lean mass from fat mass.
Does BMI work for children?
Not with the adult cut-offs on this page. Children are classified against age-specific growth references, and this calculator is intended for people aged 16 and over.
Why does my BMI say I am overweight but my waist is not?
BMI and waist circumference measure different things. BMI relates weight to height, while waist circumference approximates abdominal fat. Someone with a large frame and low body fat can have a BMI above 25 and a waist below the risk threshold. Using both together gives a better picture than either alone.