BMI Calculator

What this calculator computes

BMI scales a weight for a height so that people of different statures can be compared on one number: weight in kilograms divided by height in metres squared. Entered as 70 kg at 175 cm, that division returns 22.9 (normal weight). It is a screening ratio rather than a body-fat measure — useful because it needs only two measurements, and limited for exactly the same reason, since neither of them carries any information about how a weight is composed.

The imperial version — 703 × lb ÷ in² — is not a second rule to learn. It is the same equation with the unit conversions folded into one constant. This page does not use the shortcut: imperial entries are converted to kilograms and centimetres by the engine's own conversion helpers and then handed to the same function the metric mode calls, so the two unit modes cannot quietly disagree. Sent through that path, 154 lb at 5 ft 10 in comes back as 22.1 (normal weight).

Where that constant comes from, why the denominator is a square rather than a cube, and how far the ratio can be stretched before it stops describing anything are worked through at length in the companion guide, what BMI does and does not measure. What follows here is what is worth knowing while the form above is open.

Reading the result this page gives you

The result box shows three things: the ratio to at most one decimal place, the WHO category that ratio falls into, and the adult bands for reference. Those bands are the standard ones:

• Under 18.5 — underweight
• 18.5–24.9 — normal weight
• 25–29.9 — overweight
• 30 and above — obesity

One decimal place is a deliberate ceiling rather than a limit of the arithmetic. The division is carried out at full precision internally and rounded only on the way to the screen, because a second decimal would imply an accuracy these formulas do not have. "Copy result" copies the figure and its category as a single line of text.

Where the lines fall, and why a displayed 25 can go either way

Published tables print the bands with trailing nines — 18.5–24.9, 25–29.9 — but the comparisons the engine actually makes are against the round numbers: below 18.5, below 25, below 30, and everything else. A ratio of exactly 25, unrounded, is filed as overweight.

How far one weight can travel across those bands on height alone is easier to show than to argue about. One weight, three heights, all computed at build time by the same code the calculator runs:

• 83 kg at 165 cm → BMI 30.5 (obesity)
• 83 kg at 175 cm → BMI 27.1 (overweight)
• 83 kg at 185 cm → BMI 24.3 (normal weight)

One weight, three categories, across a 20 cm span of height. Nothing about the mass changed between those three lines; only the denominator did, and it is squared.

The wrinkle worth knowing is that the category is decided on the unrounded ratio while the figure on screen has already been rounded. Those two steps happen in that order, so a displayed number sitting on a boundary does not tell you which side of it the underlying ratio fell on. At 175 cm, 76.5 kg displays as 25 and is filed as normal weight; 76.6 kg displays as 25 and is filed as overweight. That is 0.1 kg of difference, the same figure printed twice, and two different words underneath it — which is what it looks like when a reading lands on a line drawn across a continuous ratio.

Entering your measurements

The unit radio swaps in a different set of fields rather than converting what is already typed, so a metric entry does not carry across to imperial mode. Imperial height is two fields; leaving the inches box empty is read as zero inches rather than as an error. Weight in pounds is converted to kilograms before anything else happens, which means the accepted ranges — 10 to 500 kg and 50 to 280 cm — are applied to the converted values rather than to what you typed. The two ways an entry can fail are reported in two different places: an entry that is not a number at all is flagged under the field it was typed in, while a number outside the accepted range is explained in the result box instead of a figure. An empty field is not an error at all: the result box waits. The figure recalculates on every keystroke.

Both inputs carry measurement error, and the ratio propagates both. At 178 cm, 88 kg returns 27.8 and 84 kg returns 26.5 — 1.3 of a point for 4 kg, with both readings sitting in the same band: the label held still while the number underneath it moved. Height sits in the denominator and is squared, so a proportional change in height moves the result about twice as far as the same proportional change in weight. The arithmetic has no way of knowing which of two height entries was the careful one; it divides whatever it is given.

Limits to hold in mind while reading your result

The formula has no term for composition, so one figure sits equally well on very differently composed bodies: 95 kg at 180 cm returns 29.3 (overweight) whether that mass is mostly muscle or mostly fat. The insensitivity runs both ways — muscular people read high while carrying little fat, and older adults can read in the normal band while carrying little muscle — because a category is not a description of a body, it is the bucket a ratio happened to fall into.

Two further limits are not questions of accuracy at all. BMI is not meaningful during pregnancy, and questions about weight then belong with the clinician already following it. And the adult cut-offs above do not apply to children and teens — not approximately, not with caution, not at all — because pediatric practice uses age-and-sex percentile charts instead. This calculator will still divide whatever two numbers it is given; for a child, the category it prints carries no meaning. The companion guide gives each of these limits the room it deserves.

Setting a second estimate beside it

Because the ratio holds no information about composition or distribution, the useful move is to put an independent estimate next to it rather than to interrogate this one harder:

• The body-fat calculator estimates percent fat from height plus neck and waist girths (and hip, for women) by the US Navy circumference method — different inputs, a different formula, and a standard error against reference methods such as underwater weighing commonly reported at roughly 3–4 percentage points.
• The ideal-weight calculator reports Devine, Robinson, Miller, and Hamwi side by side. They disagree with each other by design, and the width of that disagreement is itself the useful output.
• The BMR and TDEE calculators describe energy expenditure rather than shape. The BMR figure carries a band of roughly ±10% around measured resting expenditure, and the TDEE step on top of it multiplies by an activity factor that is conventional rather than precise.

None of these measure the same thing, which is exactly the point — each uses different inputs and a different formula. Everything is computed in your browser — nothing you enter is transmitted — and every worked example above is produced by the same functions at build time, so the copy and the code cannot drift apart.

Frequently asked questions

Not medical advice: this calculator provides general educational estimates from 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. Measurements are processed locally in your browser and never transmitted. See the methodology page.