On this page
BMI (Body Mass Index) is a single number that estimates body fat from your height and weight. It's the standard adult health screening metric used by the CDC and WHO. Enter your height and weight, the calculator gives you the number; the sections below explain how to interpret it and where it breaks down.
The four standard categories for adults: under 18.5 is underweight, 18.5–24.9 is normal weight, 25.0–29.9 is overweight, and 30.0 and above is obese. These bands are population-level guidelines, not individual diagnoses.
A short history of BMI
The formula was developed in the 1830s by Adolphe Quetelet, a Belgian statistician studying the average proportions of European populations. He never claimed it measured individual health, it was a population statistic designed to characterize an "average man." The label "Body Mass Index" was coined in 1972 by Ancel Keys, who studied it as the simplest formula that correlated reasonably well with body fat across large samples. Keys explicitly warned against using it for individual diagnosis.
The CDC and WHO adopted BMI as a screening metric because it's cheap, fast, and reasonably consistent at a population level, not because it's the most accurate body-composition measure available. Understanding this lineage helps explain its strengths (broad applicability, population epidemiology) and weaknesses (poor individual specificity).
Where BMI breaks down
BMI assumes that weight scales with height-squared, that all weight is metabolically equivalent, and that the population it's applied to has the same body composition as the European populations Quetelet studied. None of those assumptions holds universally:
- Muscular individuals have higher BMIs because muscle is denser than fat. An NFL running back at 5'11" / 215 lb has a BMI of 30, flagged as "obese" by the standard cutoffs while having single-digit body fat.
- Older adults lose muscle mass with age (sarcopenia), so a "normal" BMI in someone over 65 may mask elevated body fat percentage.
- Asian populations show elevated diabetes and cardiovascular risk at lower BMIs than white populations, the WHO recommends adjusted cutoffs (overweight ≥23, obese ≥27.5) for several Asian ancestries.
- Pregnant or breastfeeding women have temporary weight changes that BMI can't account for.
- Tall and very short adults are pushed toward the extremes by the
height²term, which doesn't scale physically realistically across the full height range.
For an individual reading where BMI seems off, body fat percentage, measured by DEXA scan, BodPod, or skinfold calipers, is the more accurate metric. Our body fat calculator uses the US Navy method, which estimates body fat from neck, waist, and hip circumferences without specialized equipment.
Better screening alongside BMI
A BMI number on its own is missing crucial context. Pair it with:
- Waist circumference, the standard threshold is under 35 inches for women and under 40 inches for men. Waist size correlates with visceral fat (the metabolically dangerous kind), which BMI completely ignores.
- Waist-to-hip ratio, under 0.85 for women, under 0.90 for men. A similar visceral-fat proxy.
- Blood pressure, lipid profile, and fasting glucose, these are the actual outcome measures BMI is trying to predict at the population level. In any individual case where you can measure these directly, they're far more informative than the BMI number.
Two people with identical BMIs of 27, one with a 32" waist, normal blood work, and a regular cardio habit, the other with a 42" waist, elevated triglycerides, and high fasting glucose, are in completely different health situations. The number is a starting point, not a verdict.
When to ignore your BMI
There are several situations where BMI is just the wrong tool:
- You're a trained strength or power athlete with measurable body fat under 15% (men) or 22% (women).
- You're an older adult with stable weight and good fitness markers, and your BMI puts you at the upper end of "normal" or low end of "overweight."
- You're pregnant, breastfeeding, or in the year postpartum.
- You're an adolescent under 20, adult BMI cutoffs don't apply, and pediatric BMI uses age-and-sex-adjusted percentile charts instead.
In any of these cases the "right" answer is to use a body-composition measurement directly (DEXA, BodPod) or trust the cardiovascular and metabolic indicators (blood pressure, A1c, lipid panel) over a single composite number.
Key takeaway
BMI is a population-scale screening tool, not a body-composition measurement. It treats weight as a proxy for body fat, which works on average but breaks down for muscular individuals, very tall or very short people, the elderly, and across different ethnic groups. A muscular athlete and a sedentary person can share an identical BMI while being in completely different health states. Treat the number as one input, not a verdict.
How it's calculated
The imperial formula is BMI = (weight in pounds × 703) / (height in inches)².
The metric form is the cleaner one: BMI = weight in kg / (height in meters)²,
no magic constant. The 703 factor exists only to convert pounds-and-inches into
metric-equivalent units. Both formulas produce the same number for the same person.
Where the 703 comes from
703 is just unit conversion in disguise. The metric formula is
kg / m². To convert pounds to kilograms, divide by 2.205. To convert
inches to meters, multiply by 0.0254. Working through the algebra:
BMI = (lbs / 2.205) / (in × 0.0254)² = lbs × (1/2.205) / (in² × 0.0254²)
Computing the constants: 1 / 2.205 = 0.4536 and 1 / 0.0254² = 1550.0.
Multiplying: 0.4536 × 1550.0 = 703.07. Rounded to 703, that's the
conversion factor that lets you skip the unit conversion and feed pounds
and inches directly. The number has no physical meaning beyond unit
bookkeeping.
The standard category boundaries
The CDC and WHO use these cutoffs for adults (age 20+):
| BMI range | Category |
|---|---|
| Under 18.5 | Underweight |
| 18.5–24.9 | Normal / healthy weight |
| 25.0–29.9 | Overweight |
| 30.0–34.9 | Obese class I |
| 35.0–39.9 | Obese class II |
| 40.0+ | Obese class III (severe) |
These boundaries come from large-scale population studies correlating BMI with all-cause mortality, cardiovascular disease incidence, and type-2-diabetes risk. The cutoffs are statistical thresholds where population risk noticeably elevates, not points where individual physiology meaningfully changes.
The cutoffs are also somewhat arbitrary at the boundaries. The mortality curve for BMI is U-shaped, both very low and very high BMIs correlate with elevated risk, but it's relatively flat in the 22–29 range. A BMI of 24.9 vs 25.0 doesn't represent a real biological difference; it just happens to be where the WHO drew the line.
Source: CDC adult BMI categories
Examples
-
5'9" tall, 165 lb
BMI 24.4
At 5'9" (69 inches) and 165 lb, BMI works out to 24.4, within the normal range but near the upper boundary. Gaining 8 lb would push it past 25 into the overweight band.
-
6'2" tall, 210 lb
BMI 27.0
At 6'2" (74 inches) and 210 lb, BMI is 27.0, overweight by the standard bands. For a person carrying significant muscle mass at this height, the BMI category may overstate health risk; body fat percentage gives a clearer picture in that case.
-
5'4" tall, 130 lb (typical adult woman, normal)
BMI 22.3
At 5'4" (64 inches) and 130 lb, BMI is 22.3, squarely in the normal range. There's roughly 15 pounds of margin in either direction before crossing into underweight (under ~108 lb at this height) or overweight (over ~146 lb). Most adult women in the US fall somewhere in this general band.
-
5'10" tall, 200 lb (overweight)
BMI 28.7
At 5'10" (70 inches) and 200 lb, BMI is 28.7, overweight by the standard bands but below the obesity threshold of 30. Losing about 25 lb at this height would bring BMI back into the normal range. Worth pairing with a waist circumference measurement: under 40" suggests the elevated BMI carries less metabolic risk than a 42"+ waist would.
-
5'5" tall, 195 lb (obese class I)
BMI 32.4
At 5'5" (65 inches) and 195 lb, BMI is 32.4, class I obesity. At this BMI level, population studies show meaningfully elevated risk for type 2 diabetes, hypertension, and cardiovascular disease, and the association strengthens for someone with elevated waist circumference, blood pressure, or fasting glucose. This is also a BMI level where modest weight loss (5–10% of body weight) has been shown to produce outsized improvements in metabolic markers, well before reaching a "normal" BMI.
Frequently asked questions
Is BMI accurate for everyone?
No. BMI works as a population-level screening tool but breaks down for individuals at the extremes, very muscular people (athletes, weightlifters), the elderly (who lose muscle mass), pregnant women, and across different ethnic groups. The WHO expert consultation on Asian BMI cutoffs recommends adjusted thresholds (overweight ≥23, obese ≥27.5) given elevated cardiovascular and diabetes risk at lower BMIs. For an individual reading, body fat percentage from a DEXA or BodPod scan is more accurate.
What's the difference between BMI and body fat percentage?
BMI is a ratio of weight to height-squared, it doesn't know how much of your weight is muscle, fat, bone, or water. Body fat percentage measures how much of your total mass is fat tissue specifically. A 200 lb athlete with 10% body fat and a 200 lb sedentary person with 30% body fat have the same BMI but very different metabolic profiles.
Should I be worried about a BMI in the "overweight" range?
Not necessarily. The 25–29.9 range is associated with elevated population-level risk for some conditions, but individual risk depends on body composition, muscle mass, fitness level, blood pressure, lipid profile, and family history. A BMI of 27 with low body fat, normal blood work, and regular exercise is very different from a BMI of 27 with high visceral fat and metabolic syndrome.
Is BMI the same for men and women?
The cutoffs are the same, yes. But the relationship between BMI and body fat differs, women on average have higher body fat percentages than men at the same BMI. This is one reason why some clinicians supplement BMI with waist circumference (under 35" for women, under 40" for men is the standard threshold) for a fuller picture.
Is BMI used differently for kids and teens?
Yes, and adult cutoffs do not apply to anyone under 20. Pediatric BMI is evaluated as a percentile within age-and-sex-adjusted growth charts: under the 5th percentile is "underweight," 5th–84th is "normal," 85th–94th is "overweight," 95th+ is "obese." A 10-year-old with a BMI of 19 might be normal weight (around the 65th percentile for that age), the same number in an adult would be solidly in the "normal" range too, but it means something different developmentally. The CDC publishes the percentile tables; pediatricians plot growth at well-child visits.
Should I use BMI if I'm very tall or very short?
With caution. The height² term in BMI was an empirical fit to
average European populations in the 1830s and doesn't perfectly
capture how body mass scales with height across the full range.
Very tall people (over 6'4") tend to have BMIs that read slightly
higher than their actual body composition warrants; very short
people (under 5'0") tend to read slightly lower. The deviation is
usually a point or two, not enough to flip categories on its own,
but worth knowing if you're near a boundary.
Does BMI account for muscle mass?
No, and this is its single biggest individual-level weakness. Muscle is denser than fat (about 1.06 g/mL vs 0.9 g/mL), so a muscular body weighs more for its volume. A trained strength athlete can carry an "obese" BMI of 30+ with single-digit body fat. The formula sees only weight and height, so it can't distinguish a heavy-but-lean athlete from a heavy-and-soft sedentary person. For athletes, body fat percentage measured via DEXA, BodPod, or callipers is the right tool.
What's the lowest "healthy" BMI?
The CDC/WHO floor for the normal range is 18.5. Below that is underweight, which carries its own elevated risks: weakened immune function, fertility issues, osteoporosis risk, and elevated all-cause mortality in long-term studies. The relationship between BMI and mortality is U-shaped, the bottom of the curve sits around BMI 22–25 in most large studies. Going below 18.5 isn't safer than being overweight; it's a different kind of risk.
How much weight do I need to lose to drop a BMI category?
Roughly 5–7 lb per BMI unit at average adult heights. At 5'9", each BMI unit is about 6.7 lb; at 5'4", about 5.8 lb; at 6'0", about 7.4 lb. So someone at BMI 27 (overweight) at 5'9" needs to lose roughly 14 lb to reach BMI 25 (normal). A lot of clinicians focus less on "drop to normal BMI" and more on losing 5–10% of body weight, the threshold at which most metabolic improvements (blood pressure, lipids, glucose tolerance) appear, regardless of whether the result is technically below the BMI 25 cutoff.