What BMI actually measures
Body Mass Index is a single number derived from your weight and height. The formula is weight in kilograms divided by height in meters squared: kg/m². If you live in a country that uses pounds and inches, the equivalent is 703 times weight in pounds divided by height in inches squared. Either way, the result is a number that typically falls somewhere between 15 and 45 for adults. The appeal of the number is that it is cheap, fast, and reproducible — anyone with a scale and a tape measure can compute it.
BMI was not invented as a health diagnostic. It was introduced in the 1830s by the Belgian astronomer and statistician Adolphe Quetelet, who was looking for a way to describe the "average man" in a population, not to assess any individual's health. The medical adoption came much later, largely because life insurance companies in the 20th century noticed a correlation between BMI and mortality in their policyholder data. The number became a screening tool, not a diagnosis, and that distinction still matters.
The formula and the categories
The World Health Organization defines the standard adult categories. Below 18.5 is underweight. From 18.5 to 24.9 is normal weight. From 25.0 to 29.9 is overweight. From 30.0 to 34.9 is obese class I, 35.0 to 39.9 is obese class II, and 40.0 or above is obese class III, sometimes called severe or morbid obesity. These thresholds were chosen based on epidemiological data showing that mortality risk rises sharply above a BMI of 30, and rises again above 35.
The categories are age-independent and sex-independent, which is both their strength and their weakness. A 25-year-old woman who is 165 cm tall and weighs 68 kg has a BMI of 25.0, which puts her just barely in the overweight category. A 65-year-old man with the same numbers gets the same label. Whether that label means the same thing for both people is a different question, and the answer is: not really. For older adults, a slightly higher BMI (up to about 27) is associated with better survival, partly because weight loss in older age is often a sign of illness rather than cause for celebration.
Where BMI gets it wrong
The biggest single failure of BMI is that it cannot distinguish muscle from fat. Muscle is roughly 18 percent denser than fat, so a muscular person will have a higher BMI than a less muscular person of the same height and weight. Bodybuilders, sprinters, and most professional athletes in strength sports routinely land in the overweight or even obese categories despite body fat percentages in the single digits. Dwayne Johnson at his playing weight had a BMI in the obese range. So did most of the New Zealand All Blacks. The number is technically correct; the label is misleading.
The second failure is that BMI says nothing about fat distribution, which matters more than total fat for cardiovascular risk. Visceral fat stored around the abdomen is metabolically active and strongly linked to type 2 diabetes, hypertension, and heart disease. Subcutaneous fat stored on the hips and thighs is far less dangerous. Two people with identical BMI can have very different metabolic profiles depending on where their fat sits. A simple waist circumference measurement, taken at the level of the iliac crest, captures this distinction better than BMI ever can.
The third failure is at the population edges. BMI underestimates body fat in older adults, who lose muscle mass with age, and overestimates it in young adults who are still building muscle. It underestimates body fat in people of South Asian descent, who tend to carry more visceral fat at lower BMIs, which is why several Asian health authorities recommend lower cutoffs (overweight at 23.0, obese at 27.5) for these populations.
Better measurements to consider
If you want a single number that does better than BMI, measure your waist. A waist circumference above 94 cm (about 37 inches) in men or 80 cm (about 31.5 inches) in women is associated with increased cardiovascular risk. Above 102 cm (40 inches) in men or 88 cm (35 inches) in women, the risk is substantially higher. The measurement is cheap, requires only a tape measure, and predicts outcomes better than BMI does in most studies. The waist-to-height ratio, which divides waist by height, performs even better: as a rule of thumb, keep your waist circumference below half your height.
For a more precise picture, body composition measurement separates fat mass from fat-free mass. The cheapest method is a skinfold caliper in the hands of someone trained, which gives a body fat percentage within about 3 to 4 percent of true value. Bioelectrical impedance scales, the kind sold for home use, are convenient but their accuracy is poor — hydration status alone can swing the reading by several percentage points. DEXA scans, the gold standard, cost between $50 and $150 in most cities and give you bone density, fat mass, and lean mass by region. For most people, the tape measure is plenty.
BMI for children and teens: the percentile system
BMI works differently for anyone under 20, because children are still growing and their body composition changes with age. A 10-year-old with a BMI of 20 is not classified as overweight or normal; she is classified by where her BMI falls relative to other 10-year-olds of the same sex. The CDC and WHO growth charts use percentiles: below the 5th percentile is underweight, 5th to 85th is healthy weight, 85th to 95th is overweight, and 95th or above is obese. A child at the 95th percentile has a higher BMI than 95 percent of children of the same age and sex.
The percentile system exists because a single BMI threshold would misclassify too many children. A 14-year-old boy with a BMI of 22 might be perfectly healthy if he is in the middle of a growth spurt and adding muscle, or he might be carrying excess body fat if his peers have lower BMIs. The percentile puts the number in context. Parents who try to apply adult thresholds to their children get misleading results: a BMI of 25 is overweight for an adult but may be normal for a teenage boy in the middle of puberty.
The percentile system has its own limitations. It assumes that the reference population is healthy, which is questionable when the reference population itself has high rates of childhood obesity. The CDC growth charts were updated in 2000 based on data from the 1960s through 1990s, and a child at the 85th percentile on those charts is at a higher absolute BMI today than the same percentile would have been a generation ago. For tracking an individual child over time, the trend in their own percentile matters more than any single measurement. A child whose percentile rises sharply over six months deserves attention regardless of where they started.
How to use BMI sensibly
BMI is useful as a screening tool at the population level and as a tracking tool for individuals over time. If your BMI is 33 and you have not been weighed in a year, the number tells you something. If your BMI was 22 last year and is 27 this year without an obvious explanation, the change matters more than the absolute number. Use BMI the way you use a bathroom scale: as a coarse trend indicator, not as a final verdict on your health.
If your BMI lands you in a category that feels wrong — you are visibly lean and the calculator calls you overweight — ignore the calculator and measure your waist instead. If your BMI is in the normal range but your waist circumference is above the threshold, the waist measurement is the one to act on. The two numbers together tell you more than either alone. Most importantly, treat BMI as one input among many: blood pressure, resting heart rate, fasting glucose, lipid panel, and how you feel climbing three flights of stairs all carry information that BMI does not.
A practical action plan
Start by measuring your waist at the level of your hip bones, after a normal exhale, with the tape measure parallel to the floor. Write the number down. Then compute your BMI from your weight and height. If both numbers are inside the standard ranges, your focus should shift away from weight management and toward fitness: cardiovascular capacity, strength, and mobility matter more than another decimal point on the BMI scale.
If either number is outside the range, set a target that involves a measurable change rather than a vague intention. Losing 5 to 10 percent of body weight is enough to substantially reduce cardiovascular risk in most people, and it is achievable over three to six months for almost anyone. Aim for a rate of 0.5 to 1 kg per week. Faster than that, you are losing water and lean mass, not fat. Pair the target with concrete habits: a 30-minute walk most days, two strength training sessions a week, protein at every meal, sleep of at least seven hours. Re-measure monthly, ignore the day-to-day noise, and adjust when the trend stops moving.