Is BMI Accurate? Who It Overestimates and Underestimates For

What BMI Actually Measures

BMI is not accurate at the individual level for anyone with an unusual muscle-to-fat ratio, because it cannot tell the two apart. It’s an accurate reflection of one thing only: weight scaled against height, using the formula weight in kilograms divided by height in metres squared (kg/m²). Two people can share the exact same BMI, height, and weight, and carry completely different amounts of body fat, because the formula has no way to separate a kilogram of muscle from a kilogram of fat.

That’s not a flaw in the arithmetic. It’s a limit built into what the arithmetic was ever designed to do. At a population level, averaged across thousands of people, BMI correlates reasonably well with body fat percentage and with risk of conditions like type 2 diabetes and heart disease. At an individual level, that correlation gets noisy, and for specific groups it breaks down in a predictable direction.

Where the Formula Came From

The formula behind your BMI calculator is nearly 200 years old, and it wasn’t built by a doctor. Adolphe Quetelet, a Belgian mathematician and astronomer, developed what he called the Quetelet Index in the 1830s while trying to describe the statistical “average man” across large populations, for social-science purposes, not clinical ones. He never intended it to diagnose an individual’s health.

It stayed a niche statistical tool until 1972, when physiologist Ancel Keys and colleagues published a paper comparing several weight-for-height indices and found Quetelet’s version tracked body fat about as well as more complex methods, while being far simpler to calculate. Keys renamed it the Body Mass Index. Insurance companies and health agencies adopted it over the following decades because it was cheap, fast, and needed nothing but a scale and a tape measure – not because it was the most precise option available.

Who BMI Overestimates Fat For

Muscular and highly trained people are the clearest case where BMI gets it wrong. Muscle is denser than fat, so it takes up less space per kilogram, but BMI’s formula weighs a kilogram of muscle exactly the same as a kilogram of fat. A rugby prop, a competitive bodybuilder, or a strength athlete can carry very little body fat and still land in the “overweight” or “obese” BMI band purely because of muscle mass on the scale.

This isn’t a rare edge case limited to elite athletes. Anyone doing serious resistance training for a few years, without an especially lean build to begin with, can see their BMI climb into the overweight range as muscle is added, even while their waistline shrinks and their body fat percentage drops. The body fat calculator and a waist measurement will tell that story far more accurately than the BMI number alone.

Who BMI Underestimates Fat For

Older adults are the mirror-image problem. From roughly the 40s onward, most people gradually lose muscle mass – a process called sarcopenia – even if their weight on the scale stays flat. Fat quietly replaces some of that lost muscle. The BMI number can stay exactly the same for twenty years while the body composition behind it shifts substantially toward more fat and less muscle, because BMI only ever sees the total, never the mix.

This is one reason some geriatric research has questioned whether the standard 18.5-24.9 “normal” band is the right target for people over 65, since a slightly higher BMI in that age group doesn’t always carry the same risk it does at 35. That’s an active area of research rather than a settled recommendation, and it’s a reason to treat an older adult’s “normal” BMI as a starting point for a conversation with a doctor, not a clean bill of health on its own.

Different Reference Points by Population

The standard 18.5-24.9 healthy range isn’t applied identically everywhere, because health risk doesn’t start at the same BMI for every population. The WHO Expert Consultation (Lancet, 2004) reviewed data showing that for many Asian populations, the risk of diabetes and cardiovascular disease begins climbing at a lower BMI than in the general international dataset. Rather than replacing the international categories, WHO added lower “action points” – 23.0 as a signal to watch more closely, 27.5 as an obesity-equivalent action point – specifically for public health screening in Asian populations, while keeping the standard categories as the international default.

Children and teenagers aren’t measured against the fixed adult bands at all. Growing bodies change proportion constantly, so paediatric guidance (the CDC growth charts are the commonly used reference) scores BMI-for-age as a percentile compared with other children of the same age and sex, not against a flat 18.5-24.9 cutoff.

Pregnancy is its own case. Healthy weight gain during pregnancy is expected and necessary, so standard BMI categories aren’t a meaningful or recommended way to assess a pregnant person’s weight – clinicians use pregnancy-specific weight-gain guidance instead, tied to pre-pregnancy BMI as a starting point rather than a category to hit during the pregnancy itself.

What Catches What BMI Misses

Because BMI can’t separate muscle from fat or tell you where fat is stored, a few other measurements are worth running alongside it rather than instead of it. Waist-to-height ratio – keeping your waist circumference under half your height – has outperformed BMI in some research for predicting cardiovascular risk, precisely because it captures abdominal fat directly instead of inferring it from total weight. Waist-to-hip ratio does something similar. Body fat percentage, estimated through methods like the Navy tape-measure method or bioelectrical impedance scales, answers the muscle-versus-fat question BMI structurally cannot.

None of these is perfect either – tape-measure methods have their own margin of error, and home bioelectrical impedance scales can shift several points depending on hydration. The practical approach is to treat BMI, waist measurement, and body fat percentage as three imperfect instruments that are far more useful read together than any one of them is read alone. Our guide comparing BMI, body fat, and waist-to-height ratio walks through how to weigh the three against each other.

Common Mistakes Reading a BMI Result

The most common mistake is treating a single BMI number as a diagnosis rather than a screening flag – a “normal” result doesn’t rule out excess fat (particularly in older adults), and an “overweight” result doesn’t confirm it (particularly in muscular people). The second most common mistake is comparing your BMI directly against someone else’s without accounting for age, sex, muscle mass, or ethnicity, all of which shift how much any single number actually means. The third is applying the fixed adult categories to a child, teenager, or pregnant person, none of whom the standard bands were built to score.

Frequently Asked Questions

Is BMI actually accurate?

It’s accurate at what it measures – weight relative to height – but that’s an imperfect proxy for body fat. At a population level it correlates reasonably with health risk; for an individual, especially a muscular or older person, it can be off enough to mislead.

Why does BMI overestimate body fat in muscular people?

Because BMI weighs a kilogram of muscle the same as a kilogram of fat, even though muscle is denser and healthier to carry. A lean, muscular person can weigh more per unit of height than a less muscular person with more body fat, landing in a higher BMI category despite lower actual fat.

Does BMI work the same for older adults?

Not reliably. Age-related muscle loss means an older adult can keep a stable “normal” BMI while their body fat percentage rises, because BMI only tracks total weight against height, not what that weight is made of.

Is the healthy BMI range different for Asian populations?

The international categories stay the same, but WHO’s 2004 Expert Consultation added lower action points – 23.0 and 27.5 – for public health screening in Asian populations, based on evidence that cardiovascular and diabetes risk climbs at a lower BMI in that group than in the general international dataset.

Can I use BMI during pregnancy?

Not to judge a “healthy” weight during pregnancy itself. Weight gain is expected and necessary, so standard BMI categories aren’t a recommended way to assess it – pre-pregnancy BMI is instead used as a starting point for pregnancy-specific weight-gain guidance from a clinician.

What should I use instead of BMI?

Not instead – alongside. Waist-to-height ratio, waist-to-hip ratio, and body fat percentage each capture something BMI structurally can’t, and reading two or three of them together gives a far more complete picture than any single number.

Is a “normal” BMI proof that I’m healthy?

No. BMI doesn’t account for diet quality, fitness level, blood pressure, blood sugar, or where fat is stored on the body, all of which affect health risk independently of the number on a BMI calculator.

The Bottom Line

BMI is a fast, free, population-built screening number, not a personal diagnosis. It’s worth checking, and worth taking seriously as one data point among several – just not worth treating as the final word, especially if you’re notably muscular, older, pregnant, still growing, or from a population where the standard cutoffs are known to sit in the wrong place. Pair it with a waist measurement or body fat estimate and you’ll have a far more honest read on where you actually stand.