What the number means
BMI = weight (kg) ÷ height (m)². A person who weighs 70 kg at 1.75 m: 1.75² = 3.0625, and 70 ÷ 3.0625 = 22.9. To run the maths yourself, see How to Calculate BMI by Hand, which works through metric and imperial examples step by step. The standard bands, from the World Health Organization: below 18.5 is underweight, 18.5 to 24.9 is the normal range, 25 to 29.9 is overweight, 30 and above is obesity. So 22.9 sits comfortably in the normal band.
Those bands were drawn from population data, which matters more than it sounds. They describe where risk rises across large groups, not where your personal danger starts. Keep that distinction. It explains nearly every argument about BMI.
What BMI gets right
It is cheap, instant, and needs nothing but a scale and a tape measure. Across populations, higher BMI tracks with higher risk of the diseases everyone worries about. As a first filter, a way to flag who deserves a closer look, it does its job. Public health runs on tools like this because the perfect measurement, applied to nobody, helps nobody.
Where it fails: muscle
BMI cannot see the difference between a kilogram of muscle and a kilogram of fat. Take an athlete: 90 kg at 1.80 m. 1.80² = 3.24, and 90 ÷ 3.24 = 27.8. The bands call that overweight, bordering on obesity. Look at the person and the label is absurd. This is the most common way BMI is “wrong”, and it is not a rare edge case. Anyone who lifts seriously, plays a power sport, or does physical labour can land here.
The reverse happens too. Someone with little muscle and a soft middle can sit at 23, inside the normal band, while carrying the exact belly fat the bands are supposed to warn about. The number said fine. The mirror disagrees. Researchers call this normal-weight obesity, and it is the quiet failure mode nobody posts about.
Where it fails: age and build
Muscle melts with age. Two people can share a BMI of 26 while one is a fit 30-year-old and the other is a 70-year-old who has lost a third of their muscle. Same number, very different bodies, very different risk. BMI has no age dial.
Build matters across populations as well. Research, including World Health Organization consultations on Asian populations, has suggested that health risks can rise at lower BMI values for some groups, with lower trigger points discussed than the standard 25 and 30. One global formula was never going to fit every body. Treat the bands as a starting point calibrated on averages, not a personal verdict.
What doctors actually check
When a doctor looks past BMI, the next measurement is usually the waist. Belly fat is the risky kind, and a tape measure finds it directly. A widely used rule of thumb: keep your waist under half your height. At 175 cm tall, that means a waist under 87.5 cm. It is not official WHO gospel, but many clinicians use it because it catches the normal-weight obesity case that BMI misses.
After that come the measurements BMI was always a proxy for: blood pressure, blood sugar, cholesterol and triglycerides, and sometimes a body-fat percentage via calipers or a scan. Notice what happened. The single number became five numbers. That is the real answer to “is BMI accurate”: it was never supposed to be the whole checkup.
So should you ignore your BMI?
No. Use it as a flag, not a verdict. If your BMI is 22.9, the flag is down; carry on. If it is 31, the flag is up; the correct response is a checkup, not an argument with the formula. And if you are the athlete at 27.8 who trains five days a week, you already know the number is misreading you. The failure mode to actually worry about is the quiet one: a “normal” BMI with a growing waistline.
Run your own numbers on our BMI Calculator, then read the result the way this guide suggests: as the start of a question, not the end of one.