BMI: what it actually tells you (and what it doesn't)

A 200-year-old metric still used everywhere. Where it's useful, where it falls apart.

Body Mass Index is the weight metric you'll see at every doctor's visit, every insurance form, and every fitness app. It's also one of the most criticized metrics in modern health, with good reason. It works, in a narrow sense, for a narrow purpose. The trouble starts when people use it for something else.

What BMI is

BMI is a single number derived from height and weight: 

BMI = weight(kg) / height(m)²

For US units: BMI = 703 · weight(lb) / height(in)².

The formula was published by Belgian mathematician Adolphe Quetelet in 1832 — not as a medical tool but as a way of describing the average build of a population. It got renamed "BMI" in 1972 and became the standard public-health screening tool largely because it requires only a tape measure and a scale.

The standard categories

The WHO categorizes adult BMI as:

  • Under 18.5 — underweight
  • 18.5 to 24.9 — normal weight
  • 25.0 to 29.9 — overweight
  • 30.0 to 34.9 — obesity class I
  • 35.0 to 39.9 — obesity class II
  • 40.0 and above — obesity class III

These cutoffs were set based on epidemiological studies in the 1990s linking BMI ranges to mortality risk in large populations. They are population averages. They are not your personal health.

What BMI is good for

BMI is a useful first-pass screening tool because:

  • It's cheap and consistent. Anyone can compute it in seconds. That's huge for population-level health tracking, where measuring body fat directly across millions of people is impossible.
  • It correlates with mortality at the extremes. People with BMI under 18.5 or above 35 do, on average across populations, have higher mortality risk than people in the middle of the range. The middle isn't safer than the high-overweight band by much, but the very-low and very-high bands are notably riskier.
  • It's a rough proxy for body fat in untrained adults. For sedentary people of average build, BMI moves in the same direction as actual body fat percentage. It's not precise, but it's not random either.

Where BMI falls apart

Athletes and muscular builds

Muscle is denser than fat. A heavily-muscled person can be in the "obese" BMI range while having 12% body fat. Most NFL running backs are clinically "overweight" by BMI. Body composition matters more than weight, and BMI doesn't measure body composition.

Older adults

After about age 65, the relationship between BMI and mortality flips. Slightly-overweight older adults often have better outcomes than "normal-weight" peers, partly because the extra mass is protective during illness and partly because muscle loss in aging is hidden under stable weight. BMI categories were calibrated on younger adults and don't translate cleanly.

Different ethnic groups

BMI cutoffs were set primarily on European populations. People of South Asian descent develop metabolic disease at lower BMIs; Pacific Islander populations stay metabolically healthy at higher ones. The WHO has issued different recommended cutoffs for some groups, but most clinical software still uses the universal numbers.

Fat distribution

Two people with identical BMI can have very different metabolic risk depending on where they carry fat. Visceral fat (around the organs) is much more dangerous than subcutaneous fat (under the skin). BMI sees neither.

Children and teens

BMI in children isn't compared against fixed cutoffs but against age-and-sex percentile charts. The adult formula is the same, but the interpretation is completely different.

Better metrics, when you have them

If BMI puts you near a category boundary or in a band that doesn't match how you look or feel, these add real information:

  • Waist circumference. Measured at the navel. Above 40 inches (102 cm) for men or 35 inches (88 cm) for women is a stronger predictor of cardiovascular risk than BMI.
  • Waist-to-height ratio. Waist measurement divided by height. Below 0.5 is generally good. Simple, but captures fat distribution in a way BMI cannot.
  • Body fat percentage. Measured via DEXA scan (most accurate), bioelectrical impedance (cheap, consumer scales), or skinfold calipers (used in fitness contexts). For health, "athletic" ranges are roughly 10–20% for men, 18–28% for women.
  • Lean body mass. Total weight minus fat mass. Often more useful than BMI for tracking changes during a training program.

How to actually use BMI

  • Treat it as a rough screen, not a diagnosis.
  • If your BMI is normal and you have no other risk markers, don't worry about it.
  • If your BMI is in the overweight or obese band, the BMI itself isn't telling you something new — body fat percentage, waist size, blood pressure, and lipid panel will tell you whether it actually matters.
  • If you're an athlete or a heavily-trained adult, it almost certainly overstates your risk. Use body composition instead.
  • Track the trend, not the number. Going from BMI 32 to BMI 28 is meaningful regardless of what category you started in.

Calculate yours in our BMI calculator, and compare with our body fat estimator and lean mass calculator to see the bigger picture.