BMI Calculator

BMI = weight (kg) / height (m)^2 was popularized by 19th-century Belgian statistician Adolphe Quetelet to describe populations, not individuals. The formula contains zero information about what the weight is made of. Skeletal muscle has a density of roughly 1.06 g/cm^3, while adipose tissue (fat) has a density of about 0.9 g/cm^3 — that 18% density gap means a lean, muscular body weighs more for the same volume than a fatter body. So a 95 kg, 180 cm rugby player at 12% body fat scores BMI 29.3 (CDC "overweight, near obese"), while a 95 kg, 180 cm sedentary office worker at 32% body fat scores the exact same 29.3. The number is identical. The cardiovascular risk profile is opposite. Direct body composition methods exist and are dramatically more informative. DEXA (dual-energy X-ray absorptiometry) is considered the clinical gold standard, with typical errors of ±1–2 percentage points of body fat — it scans bone, lean mass, and fat separately. BIA (bioelectrical impedance analysis), the technology used by smart scales like the Withings Body+ or InBody, sends a tiny current through the body and estimates fat from electrical resistance; accuracy is sensitive to hydration status (drink a liter of water and your reading shifts) but trends across weeks are useful. Hydrostatic (underwater) weighing exploits the density gap between fat and lean tissue directly via Archimedes' principle and is highly accurate but inconvenient. Skinfold calipers, when used by a trained technician across 7 sites (Jackson-Pollock protocol), come within ±3% of DEXA at a fraction of the cost. Waist-to-hip ratio (WHR) and waist-to-height ratio (WHtR) are the two simplest BMI complements that anyone can measure with a tape. The World Health Organization defines abdominal obesity as WHR ≥ 0.90 in men and ≥ 0.85 in women. WHtR ≥ 0.5 — meaning your waist is at least half your height — is the threshold associated with elevated metabolic risk in multiple meta-analyses, including Ashwell et al. (2012, BMJ Open). Visceral fat, the deep abdominal fat that surrounds organs, is metabolically active and far more strongly linked to type 2 diabetes and cardiovascular disease than subcutaneous fat — and BMI does not see it. The bottom line: use BMI as a population-level screening tool and a starting flag, never as a diagnosis. If your BMI flags you as overweight but your waist is small, you carry visible muscle, and your blood markers are normal, BMI is wrong about you specifically. If your BMI is "normal" but you have a large waist, low muscle mass, and elevated blood glucose ("normal-weight obesity" or TOFI — thin-outside-fat-inside), BMI is also wrong about you. This is informational only; consult a healthcare professional for assessment of individual risk.

// Two bodies, identical BMI, opposite risk profiles
const player  = { kg: 95, m: 1.80, bodyFatPct: 12 }; // muscular athlete
const sitter  = { kg: 95, m: 1.80, bodyFatPct: 32 }; // sedentary worker
const bmi = (b) => b.kg / (b.m * b.m);
bmi(player); // 29.32 -> CDC "overweight"
bmi(sitter); // 29.32 -> CDC "overweight" (identical!)

// Waist-to-Height Ratio: a better single-tape signal
function whtr(waistCm, heightCm) { return waistCm / heightCm; }
whtr(95, 180);  // 0.527 -> elevated metabolic risk (Ashwell threshold 0.5)
whtr(78, 180);  // 0.433 -> low risk

// Body fat density vs muscle density (kg per liter)
const muscleDensity = 1.06;  // g/cm^3
const fatDensity    = 0.90;  // g/cm^3
// Same volume, muscle weighs ~18% more

The familiar WHO cutoffs — underweight under 18.5, normal 18.5–24.9, overweight 25–29.9, obese 30+ — were derived from cohorts that were predominantly white European and North American. They do not apply uniformly across ethnicities, and pretending they do produces serious clinical errors. In 2004, after the WHO Expert Consultation reviewed evidence from East Asian, South Asian, and Pacific populations, a separate set of "Asian-specific cutoffs" was published: overweight starts at BMI 23, and obese starts at BMI 27.5 (WHO 2004, Lancet 363:157–163). The reason is metabolic, not cosmetic — for a given BMI, Asian populations tend to carry higher percentage body fat and more visceral fat, particularly around the abdomen, and develop type 2 diabetes and cardiovascular disease at substantially lower BMI than European-descent populations. The Korean Society for the Study of Obesity (KSSO) explicitly uses BMI 23 as the overweight threshold and 25 as the obese threshold, even more conservative than the WHO Asian cutoffs. Black populations face the opposite calibration problem in the other direction. African and African-American adults at the same BMI tend to have higher lean mass and lower visceral fat than white peers, meaning the standard 25 and 30 thresholds may slightly over-flag risk. South Asian populations (Indian, Pakistani, Bangladeshi) carry the highest visceral fat per BMI of any major group studied — UK NICE guidelines (2013, updated 2023) recommend BMI 23 overweight / 27.5 obese for South Asian, Chinese, other Asian, Middle Eastern, Black African, and African-Caribbean adults living in the UK. Polynesian and Pacific Islander populations, by contrast, often have higher lean mass and bone density, making 30 a less reliable obesity flag. Practical implication: the result this calculator shows uses the standard WHO cutoffs. If you are of East or South Asian descent, mentally subtract roughly 2 BMI points from each threshold. If your BMI is 24 and you are Korean, Chinese, or Indian, your risk profile is closer to what BMI 26 implies for someone of European ancestry. This is not a reason to panic — it is a reason to look at additional markers (waist circumference, fasting glucose, HbA1c, lipid panel) and to discuss with a doctor familiar with ethnicity-specific risk thresholds. Educational information only; not medical advice.

// Standard WHO vs WHO-Asian vs Korean (KSSO) cutoffs
function classify(bmi, scheme = 'who') {
  const cuts = {
    who:     { over: 25.0, obese: 30.0 },  // WHO standard
    asian:   { over: 23.0, obese: 27.5 },  // WHO 2004 Asian
    korean:  { over: 23.0, obese: 25.0 }   // KSSO (more conservative)
  }[scheme];
  if (bmi < 18.5)            return 'underweight';
  if (bmi < cuts.over)       return 'normal';
  if (bmi < cuts.obese)      return 'overweight';
  return 'obese';
}

const myBmi = 24.0;
classify(myBmi, 'who');    // 'normal'
classify(myBmi, 'asian');  // 'overweight'
classify(myBmi, 'korean'); // 'overweight'

Adult BMI categories are absolute — 25 means overweight whether you are 25 or 75 years old. Childhood BMI is different in a fundamental way: a 6-year-old's body composition is not even loosely comparable to a 14-year-old's, so the same BMI number means dramatically different things at different ages and between sexes. The CDC and WHO therefore use sex-specific BMI-for-age percentiles for everyone aged 2–19, not numeric categories. The CDC growth charts (cdc.gov/growthcharts) plot BMI against age in months, separately for boys and girls, with reference curves drawn at the 5th, 10th, 25th, 50th, 75th, 85th, 90th, 95th, and 97th percentiles. The official categorization: underweight is BMI below the 5th percentile, healthy weight is 5th to under 85th, overweight is 85th to under 95th, and obesity is 95th percentile or above. Severe obesity in children is often defined as ≥120% of the 95th percentile or ≥35 BMI absolute (whichever is lower). These cutoffs were chosen so that around 5% of the reference population would fall into each tail, defining "outlier" statistically rather than mechanistically. The reference data matters. CDC charts are based on US NHANES survey data from 1963–1980 and 1988–1994 (the agency deliberately used pre-obesity-epidemic data so the curves represent a historically healthier baseline). WHO charts, by contrast, use a multi-country sample of breastfed infants and children raised under "optimal" conditions, and tend to flag a slightly different population as underweight in the youngest ages. Pediatricians in the US typically follow CDC for ages 2+ and WHO for under 2. The exact percentile is normally reported by the pediatric EHR, computed via the LMS method (Cole and Green 1992) which transforms skewed BMI distributions into a normal distribution using three age- and sex-specific parameters: lambda (skewness), mu (median), sigma (coefficient of variation). Practical bottom line for parents using this calculator: the BMI number this tool shows is the same calculation, but interpreting it for a child requires looking at percentile curves, not the adult color bar. A 9-year-old girl with BMI 21 might be at the 92nd percentile (overweight), while a 14-year-old girl with the exact same BMI 21 might sit at the 65th percentile (healthy). Always defer to the percentile curve printed on the CDC or WHO chart, or — better — let the pediatrician interpret it. This is informational, not diagnostic; for any concern about a child's growth, see a pediatrician or family physician.

// Same BMI, very different percentile depending on age and sex.
// Real percentile lookup uses CDC/WHO LMS tables; this is illustrative.
const child = { ageMonths: 108, sex: 'F', bmi: 21.0 }; // 9-year-old girl
// CDC 2000 LMS table lookup -> roughly 92nd percentile (overweight)

const teen  = { ageMonths: 168, sex: 'F', bmi: 21.0 }; // 14-year-old girl
// CDC 2000 LMS table lookup -> roughly 65th percentile (healthy)

// Categories used for children (CDC):
//   < 5th  percentile  : underweight
//   5-84.9 percentile  : healthy weight
//   85-94.9 percentile : overweight
//   >= 95  percentile  : obesity
//   >= 120% of 95th OR >=35 BMI : severe obesity