How Accurate Is BMI? The Screening Tool's Limits

For large groups of people, BMI tracks direct measures of body fat well enough to steer public-health decisions, which is exactly why the CDC and WHO still use it. It needs nothing but height and weight, it's free, and its categories have been studied in millions of people. That kind of evidence base isn't nothing.

For an average-build adult, a BMI in the healthy range is genuinely reassuring, and a BMI above 30 is genuinely informative. The trouble starts at the edges: very muscular bodies, very lean bodies, older bodies, shorter bodies. That's where the number starts lying.

BMI can't see what your weight is made of. A strength athlete at 8% body fat can score 28 and land in the 'overweight' bucket, because muscle is dense and the formula doesn't care. Meanwhile an older adult can sit at BMI 23 while carrying 30% body fat after years of muscle loss. Researchers have a name for that second profile: normal-weight obesity. It's common, and it carries real metabolic risk.

Distribution is the other blind spot. Fat around the abdomen drives far more cardiometabolic risk than the same fat on hips and thighs, and BMI is completely blind to where your weight sits. Ethnicity matters too: some Asian populations show elevated risk at BMI cut-offs that would read as normal on Western charts.

If BMI feels wrong for your body, verify it with a second measure instead of arguing with the scale. Waist-to-height ratio (keep it at 0.50 or below) tells you where the fat is. Body-fat percentage, by Navy tape or skinfold, tells you what the weight is made of. And if you want the ground truth, DEXA is the standard, though you'll need a clinic visit.

The practical take is simple: use BMI as a first filter, not a verdict. When BMI and body composition disagree, trust the measure that actually looks at your body.