Body mass index is weight divided by height squared.
metric BMI = kg ÷ m²
US units BMI = 703 × lb ÷ in²
The 703 in the second form is simply the unit conversion; both give the same answer.
The standard adult categories
| BMI | Category |
|---|---|
| Below 18.5 | Underweight |
| 18.5 – 24.9 | Normal weight |
| 25.0 – 29.9 | Overweight |
| 30.0 and above | Obesity |
Obesity is often subdivided: class I 30–34.9, class II 35–39.9, class III 40 and above.
These bands apply to adults. For children and teenagers BMI is interpreted as a percentile for age and sex rather than against fixed numbers, because body composition changes throughout growth.
What it was built for
BMI was developed to describe populations, not to assess individuals. Across a large group it correlates usefully with body fat and with the health outcomes associated with it, and it is cheap to collect — two measurements, no equipment.
That is its strength and the source of every one of its limitations. A ratio of two numbers cannot know what the weight is made of.
Four situations where it misleads
Muscular people. Muscle is denser than fat. A well-trained athlete frequently lands in the overweight or obese category with low body fat. The number is arithmetically correct and clinically meaningless.
Older adults. Muscle mass declines with age while fat can increase, so an unchanged BMI over decades can conceal a substantial change in composition. BMI also under-identifies risk in this group for that reason.
Fat distribution. BMI says nothing about where fat is carried, and visceral fat around the abdomen carries considerably more metabolic risk than the same mass on hips and thighs. Two people at BMI 28 can have quite different risk profiles.
Ethnicity. The risk associated with a given BMI differs between populations. Several health bodies recommend lower thresholds for people of South Asian, Chinese and some other Asian backgrounds, where metabolic risk appears at a lower BMI than the standard bands suggest.
Measures that add what BMI misses
Waist circumference captures central fat directly and is a better single predictor of metabolic risk than BMI. Commonly cited action thresholds are around 40 inches for men and 35 inches for women, with lower figures recommended for some populations.
Waist-to-height ratio is simpler still: keeping waist circumference under half your height is a widely used rule of thumb, and it needs no tables.
Body fat percentage measures composition directly, though home scales that estimate it by bioimpedance are strongly affected by hydration and should be read as a trend rather than a value.
None of these replaces BMI so much as complete it. Used together — BMI plus waist measurement — they describe considerably more than either alone.
Using it sensibly
BMI is a reasonable starting point and a poor conclusion. It is most useful for tracking change in yourself over time, where the composition confound is smallest, and least useful for comparing two individuals.
If the number sits in a category that surprises you given how you look and how you feel, that is exactly the situation in which the other measures are worth taking, and in which a conversation with a clinician is more informative than a recalculation.
Worked examples
A person 5 feet 10 inches tall weighing 180 pounds: 70 inches squared is 4,900, and 703 × 180 ÷ 4,900 gives a BMI of about 25.8 — just inside the overweight band.
The same height at 165 pounds gives about 23.7, comfortably in the normal band. Fifteen pounds moves the category, which is worth knowing before treating a category boundary as meaningful. The bands are round numbers chosen for convenience, and nothing physiological happens at 25.0.
In metric, someone 1.78 m and 82 kg: 1.78 squared is 3.17, and 82 ÷ 3.17 gives about 25.9. The same arithmetic, no conversion factor.
Why the categories have moved
The current thresholds have not always been the thresholds. The cut-off between normal and overweight was lowered in the United States in 1998, which moved millions of people into a different category overnight without anyone gaining weight.
That history is worth knowing because it makes the point plainly: the bands are administrative conventions layered on a continuous measurement, chosen to be useful across populations. They are not thresholds where risk switches on.
Risk associated with body composition rises gradually and continuously, and it interacts with blood pressure, blood sugar, lipids, fitness and family history — none of which appear in the formula.