What BMI Does — and Does Not — Measure

Body mass index is one division problem: a weight divided by the square of a height. Nearly everything people argue about when they argue about BMI follows from how little that division is allowed to know. This guide works the arithmetic in both metric and imperial form, explains why the denominator is squared, sets out the WHO adult cut-offs, and spends most of its length on the documented blind spots. Every figure below is computed at build time by the same engine that runs the BMI calculator.

The ratio, and the two forms it comes in

In metric units BMI is weight in kilograms divided by height in metres squared: kg ÷ m². The imperial form looks like a different formula and is not — 703 × lb ÷ in² is the same operation with the unit conversion folded into a constant. Read from this site's own converters, a pound is 0.45359237 kg and an inch is 0.0254 m, so converting first means multiplying the weight by the pound-to-kilogram factor and dividing by the square of the inch-to-metre factor; that quotient is 703.0696. The familiar imperial shortcut is the metric formula with the conversion already done and rounded to three digits.

This site converts rather than shortcuts: imperial input passes through the engine's own pound-to-kilogram and inch-to-centimetre converters and then into the one kg ÷ m² division every other page uses, so there is a single BMI implementation here instead of two that could quietly disagree. Worked through, 5 ft 10 in and 174 lb convert to 177.8 cm and 78.9 kg, and at the site's display precision of at most one decimal place the ratio reads 25. That rounding hides a wrinkle worth returning to below.

Why the denominator is height squared

A weight means nothing without a height beside it, so an index has to scale one by the other, and the exponent is a modelling choice rather than a law of anatomy. Squaring is the convention this formula uses, and the WHO adult cut-offs are applied to whatever that one ratio returns, at every adult height.

That choice buys comparability, not realism. The numerator is a mass and the denominator is a length squared — two quantities that do not scale together. The exponent two is a compromise rather than a claim: one fixed convention, applied at every adult height.

The fixed exponent has a quieter consequence for anyone tracking the number: the same gain or loss of weight moves BMI less at a greater height and more at a shorter one, because it is divided by a larger or smaller denominator.

The adult cut-offs, and one weight at three heights

The WHO adult cut-offs, as implemented here, are: under 18.5 underweight, 18.5–24.9 normal weight, 25–29.9 overweight, and 30 and above obesity. The engine assigns the category from the unrounded ratio and rounds the displayed figure only afterwards — an order of operations that matters more than it sounds like it should.

The clearest demonstration of what the index is doing is to hold the weight still and move the height. A single 80 kg reads 31.3 at 160 cm (obesity), 26.1 at 175 cm (overweight), and 22.2 at 190 cm (normal weight). Three heights, three bands, one unchanged mass — the index working exactly as designed, and the cleanest available proof that the number describes a ratio rather than a body.

What the ratio cannot see

BMI takes two numbers. Anything absent from those two numbers is invisible to it — not down-weighted, not approximated, simply not in the calculation.

Muscle and fat weigh the same to a ratio

A kilogram of muscle and a kilogram of fat are identical to a division that never asks which is which, so muscular people can read high while carrying little fat. The US Navy circumference method makes the gap visible because it takes different inputs entirely — height plus neck and waist girths (and hip, for women), converted to a body density and then to a percentage. It never asks for a weight at all. At 175 cm, a 41 cm neck with an 82 cm waist estimates 12% body fat; a 39 cm neck with a 100 cm waist estimates 26.7%. Neither girth is one of BMI's two inputs, so nothing in the ratio distinguishes an 82 cm waist from a 100 cm one at the same weight and height.

That method is an estimate too: against reference methods such as underwater weighing its standard error is commonly reported at roughly 3–4 percentage points, and tape position and tension alone can move a reading by a point or more. But it estimates something BMI does not estimate at all — the body-fat calculator runs it and states those limits beside the result.

Low muscle inside a normal reading

The blind spot runs the other way as well. Older adults can read squarely in the normal band while carrying little muscle, because the ratio notices only that a weight has stayed where it was. What that weight is made of is not one of its inputs, so a change in composition with no change in mass moves BMI not at all.

Pregnancy, children, and teenagers

BMI is not meaningful during pregnancy, and the adult cut-offs do not apply to children and teens at all — pediatric practice uses age-and-sex percentile charts, which compare a young person against others of the same age and sex rather than against fixed numbers. These are not cases where the index is merely less accurate; they are cases where the adult bands are the wrong instrument, and running the arithmetic anyway produces a number with no category behind it.

Complementary signals, not competing ones

Waist measurements and body-fat estimates are often framed as rivals to BMI. They are better understood as different questions with different inputs. BMI asks how a weight compares with a height. The circumference method asks how girths at particular sites compare with height. The ideal-weight formulas ask a third question and answer it four incompatible ways: at 175 cm the male versions return 70.5 kg (Devine), 68.9 kg (Robinson), 68.7 kg (Miller), and 72 kg (Hamwi). That spread is deliberate — the four rules originated decades ago in clinical contexts, the Devine formula having been proposed for drug dosing — and none of them consider body composition either.

Reading them together is not the same as averaging them. Each is a low-resolution view from a different angle, and the informative part is often the disagreement — a BMI in one band alongside a circumference estimate pointing elsewhere is the kind of discrepancy a qualified professional is equipped to interpret, and a calculator is not.

A common misreading: the figure and the band

Return to the imperial example. 5 ft 10 in and 174 lb display as 25, and the engine files that same reading under normal weight. That is not a bug: the category is decided on the full unrounded ratio, which sits just under the cut-off, and rounding happens only on the way to the screen.

Reading a trend rather than a reading

Both inputs carry measurement error — which scale, which time, shoes on or off — and the ratio propagates both. A single BMI is two measurements taken on one occasion, divided. A series of them, taken the same way at the same time of day, is a slope rather than a point, and a slope is read from the differences between readings rather than from any single one.

None of that promotes a screening ratio into a diagnosis; it only keeps the number honest about its own resolution, which is the most that can be asked of two measurements and a division. If the real question is about changing weight rather than describing it, BMI is the wrong instrument — an index, not an energy model. The guide to calorie needs covers the estimates that do address intake and expenditure.

Frequently asked questions

Not medical advice: this guide provides general educational information about published population formulas. It cannot account for your individual health context and must not be used to diagnose or treat any condition — consult a qualified healthcare professional. Calculations run locally in your browser and are never transmitted. See the methodology page.