Body Mass Index is weight in kilograms divided by height in metres squared. That is the whole formula, and its simplicity explains both why it is used everywhere and why it is so often misread. It was devised in the nineteenth century to describe populations, not to diagnose individuals, and every limitation below follows from that.
The standard classification
| BMI | Category |
|---|---|
| below 18.5 | underweight |
| 18.5 – 24.9 | normal weight |
| 25.0 – 29.9 | overweight |
| 30.0 – 34.9 | obesity, class I |
| 35.0 – 39.9 | obesity, class II |
| 40.0 and above | obesity, class III |
The underweight band is sometimes subdivided as well — mild thinness from 17.0 to 18.49, moderate from 16.0 to 16.99, severe below 16.0 — though the single cut-off at 18.5 is what most charts show.
The Asian cut-offs
| Category | Standard | Asian cut-offs |
|---|---|---|
| normal | 18.5 – 24.9 | 18.5 – 22.9 |
| overweight | 25.0 – 29.9 | 23.0 – 27.4 |
| obese | 30.0 and above | 27.5 and above |
These lower thresholds are used by several national health bodies across South and East Asia, on the evidence that cardiovascular and diabetes risk rises at a lower BMI in these populations. A figure of 24 is "normal" on one chart and "overweight" on the other, and which chart applies is a question the number itself cannot answer.
The thresholds were derived largely from European populations, and that is the point of the second table. A measure calibrated on one group and applied to everyone will misclassify — which is a statement about the calibration, not about the people. The same caution applies in the other direction to populations with higher average muscle mass.
What the number cannot see
| Factor | Why it matters |
|---|---|
| muscle versus fat | BMI counts total mass; the two are not equivalent for health |
| where fat sits | abdominal fat carries different risk from the same mass elsewhere |
| age | body composition shifts with age at constant weight |
| pregnancy | BMI is not a meaningful measure during it |
| children | needs age-and-sex percentiles, not adult bands |
Waist circumference and waist-to-height ratio capture the second row, which is why clinicians often ask for them alongside BMI rather than instead of it. A ratio above 0.5 — waist more than half your height — is a commonly used flag, and it takes a tape measure rather than a formula.
Using it sensibly
As one rough input among several, BMI is useful: it is free, it takes two numbers you already know, and at a population level it tracks risk well. As a verdict on an individual's health it is not, and reading a single figure as one is the mistake the measure invites. If your number sits near a boundary, the boundary is doing more work than the evidence supports — and a conversation with a healthcare professional will tell you more than recalculating it will.
Work out your BMI offline
The free BMI Calculator handles metric and imperial, in twenty languages, without sending anything anywhere.
Open the dictionary appFrequently asked questions
What are the BMI categories?
Below 18.5 is underweight, 18.5 to 24.9 is normal, 25.0 to 29.9 is overweight, and 30.0 or above is obese. Obesity is further split into class I (30.0-34.9), class II (35.0-39.9) and class III (40.0 and above).
Why are there different cut-offs for Asian populations?
Because associated health risk appears at a lower BMI. The WHO discussed lower action points for Asian populations, and several national bodies use 23.0 as the overweight threshold and 27.5 for obesity rather than 25.0 and 30.0.
Is BMI accurate for athletes?
No. BMI does not distinguish muscle from fat, so a rower and a sedentary person of the same height and weight score identically, and well-muscled athletes are routinely classified as overweight by a measure that was never designed to assess them.