Body Mass Index (BMI) is a screening measure that relates weight to height: BMI = weight (kg) ÷ height² (m²). The World Health Organization classifies adults into five categories — underweight (<18.5), normal (18.5–24.9), overweight (25–29.9), obese class I (30–34.9), class II (35–39.9), and class III (≥40). BMI is a population-level screening tool, not a diagnostic measurement of body fat; interpretation requires knowing a person's age, ethnicity, muscle mass, and waist-to-hip ratio. This calculator supports metric and imperial units, returns both the BMI number and the WHO category, and shows the healthy weight range for your height.
Reviewed: April 23, 2026 · Author: Naveen P N, Founder — AI Calculator · Verified against: World Health Organization BMI classification, CDC adult BMI guidance, NIH NHLBI BMI tables.
Medical disclaimer. BMI is a screening tool, not medical advice. It does not diagnose obesity, assess health risk on its own, or replace consultation with a physician, registered dietitian, or qualified healthcare professional. See our disclaimer for full details.
The BMI formula — metric and imperial
The Body Mass Index formula was originally published by Belgian astronomer and statistician Adolphe Quetelet in 1832 as the "Quetelet index." The WHO adopted and renamed it in 1972 as a simple, cheap, and population-scale measure of weight relative to height. The calculation is deliberately simple so it can be done without equipment more sophisticated than a scale and a tape measure.
Why the "703" conversion factor? It converts pounds-per-inches-squared into kilograms-per-meter-squared. Specifically, 1 kg = 2.20462 lb and 1 m = 39.3701 in, so the unit conversion factor is (39.3701)² ÷ 2.20462 ≈ 703.07. Rounding to 703 produces at most a 0.1% error in the final BMI, which is immaterial given that BMI itself has a much larger inherent uncertainty.
BMI is a ratio, not a linear measure. This is why doubling someone's height while keeping weight constant quarters their BMI, not halves it. The square-of-height denominator means BMI scales sensibly across human heights — an approximation that works reasonably well between about 1.4 m and 2.0 m but breaks down at extremes.
Worked example 1 — an average adult (metric, simple case)
Scenario: A 34-year-old woman weighs 68 kg and stands 1.65 m tall. What is her BMI and which WHO category does she fall into?
Step 1 — square the height in meters:
Step 2 — divide weight by height squared:
Step 3 — interpret against the WHO table. 24.98 is at the very top of the "normal" range (18.5–24.9). Mathematically she's one rounding step away from "overweight" (25.0–29.9). For counselling or risk-communication purposes, this should be reported as "top of normal range" rather than "normal" full-stop — it matters for her insurance, her doctor's advice, and her sense of context.
Sanity check — healthy weight range at this height:
Max: 24.9 × 2.7225 = 67.8 kg
She is 0.2 kg above the upper healthy-range boundary. This framing (absolute kilograms away from range) is often more actionable than the abstract BMI number itself.
Worked example 2 — a muscular athlete (imperial, BMI's biggest blind spot)
Scenario: A 28-year-old professional rugby player weighs 215 lb at a height of 5 feet 11 inches. His body-fat percentage, measured by DEXA scan, is 11% — well within athletic/lean range. What BMI does the formula give, and what does it mean?
Step 1 — convert height to total inches:
Step 2 — apply the imperial formula:
Step 3 — classify. 29.98 is in the "overweight" WHO category and one decimal away from "obese class I" (≥30). But DEXA says he has 11% body fat. Something is wrong.
What went wrong: nothing — the BMI formula is doing exactly what it's designed to do. BMI cannot distinguish muscle mass from fat mass. A rugby player, powerlifter, NFL linebacker, or gymnast who carries 10–30 lb more muscle than a sedentary person of the same height will land in the "overweight" or "obese" BMI range despite being objectively healthier. This is the canonical example of why BMI is a screening tool and not a diagnosis.
Correct interpretation for this athlete: report BMI with a caveat — "BMI 30.0, but body-fat % and waist-to-hip ratio within athletic range; BMI not a reliable indicator of adiposity for this individual." Primary care providers and insurance medical officers are trained to make this distinction; unfortunately, corporate wellness programs and automated health-screening forms often are not.
Complementary measurements to use alongside BMI in these cases: body-fat percentage (see our body-fat calculator), waist circumference (<94 cm for men, <80 cm for women per WHO), waist-to-hip ratio, and if available, DEXA or bioimpedance measurement.
Worked example 3 — an older adult in the underweight range (real-world clinical scenario)
Scenario: A 78-year-old woman, 5 ft 4 in tall, weighs 105 lb. She reports unintentional 8 lb weight loss over the past 6 months. What is her BMI, and what is the clinical concern?
Step 1 — height in inches: 5 ft 4 in = 64 in.
Step 2 — BMI:
Step 3 — classification and clinical context. 18.02 is below the WHO "normal" threshold of 18.5, placing her in the "underweight" category. For an older adult, this is a clinical flag rather than an innocuous finding:
- Elderly underweight is riskier than elderly overweight. Multiple population studies show all-cause mortality U-curves shift rightward with age — the BMI associated with lowest mortality is 23–26 for adults 65+, not 18.5–24.9.
- Unintentional weight loss > 5% in 6 months in an older adult warrants workup for cancer, heart failure, depression, dementia-related feeding difficulty, malabsorption, and medication side effects.
- Sarcopenia (age-related muscle loss) is often hidden by BMI. Two older women at BMI 20 can have very different muscle mass; the one with sarcopenia has much higher fall and mortality risk.
The correct clinical report would be: "BMI 18.0 (underweight category) with unintentional 7.5% weight loss over 6 months. Recommend dietary assessment, screening labs, and evaluation for underlying medical cause."
Common mistakes — five errors that misinterpret a BMI result
- Treating BMI as a measure of body fat. BMI is a measure of weight relative to height — it correlates with body fat at the population level but can badly misclassify individuals with atypical body composition (athletes, elderly, amputees, ectomorphs, endomorphs).
- Using adult BMI thresholds for children. For anyone under 20, BMI must be interpreted using age- and sex-adjusted percentile charts (CDC or WHO) rather than fixed thresholds. A "BMI of 22" means very different things for a 10-year-old girl (around 85th percentile, approaching overweight) vs a 17-year-old girl (around 40th percentile, normal).
- Using the 18.5–24.9 range for all ethnicities. The WHO has published separate thresholds for Asian populations (overweight begins at 23, not 25) because Asian body composition at a given BMI carries higher metabolic risk. Using non-adjusted thresholds can mean missing prediabetes signals in South and East Asian patients.
- Ignoring the difference between screening and diagnosis. A single BMI reading above 30 does not "diagnose obesity" the way a glucose reading above 126 mg/dL diagnoses diabetes. Clinical obesity diagnosis requires BMI plus assessment of body composition, metabolic status, and weight-related health conditions.
- Reading too much into small BMI changes. BMI has measurement noise: a full glass of water weighs 0.25 kg, clothing weighs 1–2 kg, time-of-day fluctuation is routinely 1–2 kg. A 0.3-point week-over-week BMI change is noise, not signal. Look at 3–6 month trends measured under consistent conditions (same time of day, same scale, same clothing state).
When BMI is the wrong tool — five populations where it misleads
- Highly muscular adults (bodybuilders, rugby, American football, Olympic weightlifting). Use body-fat percentage or DEXA.
- Pregnant women. Weight gain during pregnancy is expected and healthy; standard BMI thresholds don't apply. Use pre-pregnancy BMI as a baseline only.
- Children and adolescents (under 20). Use BMI-for-age percentile from the CDC growth charts.
- Elderly adults (65+). Shifted thresholds — normal range is 23–28 rather than 18.5–24.9. Unintentional loss of BMI is a warning sign, not necessarily good news.
- Amputees or people with significant limb-length or bone-density differences. The formula's implicit assumption of typical human body proportions fails; clinical judgement and alternative measures are needed.
WHO BMI classification reference table (adults, 20 and over)
| Category | BMI range (kg/m²) | Associated risk (general adult) |
|---|---|---|
| Severe underweight | <16.0 | High risk (malnutrition, immune, fertility) |
| Moderate underweight | 16.0–16.9 | Moderate risk |
| Mild underweight | 17.0–18.4 | Mild risk |
| Normal range | 18.5–24.9 | Average population risk |
| Overweight (pre-obese) | 25.0–29.9 | Increased risk |
| Obese class I | 30.0–34.9 | Moderately increased |
| Obese class II | 35.0–39.9 | Severely increased |
| Obese class III (morbid) | ≥40.0 | Very severely increased |
Asian-population thresholds (WHO Western Pacific): overweight begins at 23, obese class I at 27.5. This shift reflects higher body-fat-at-BMI and higher diabetes/cardiovascular risk-at-BMI observed in East and South Asian populations.
Where BMI is used — 7 real-world contexts
- Primary-care health screening. Routine annual physicals in most countries log height, weight, and BMI as the first-line weight-status signal, flagging patients for further evaluation.
- Epidemiology and public health statistics. The prevalence of overweight and obesity in any population is reported via BMI. WHO's global status reports on NCDs (non-communicable diseases) anchor obesity metrics to BMI category prevalence.
- Life and health insurance underwriting. Many insurers apply premium adjustments at BMI thresholds (often >30 and >35), though progressive insurers supplement with waist measurements and lab data.
- Pre-operative assessment. Bariatric surgery candidates are typically screened using BMI ≥ 35 with comorbidity or BMI ≥ 40 without. Elective surgeries may have BMI caps for perioperative risk.
- Pediatric growth tracking. BMI-for-age percentile is plotted alongside height and weight percentiles from infancy through adolescence on every well-child visit.
- Sports team and military fitness qualification. Some service branches use BMI as an initial screen; those who fail then progress to body-fat measurement before disqualification. See our body-fat calculator using the US Navy method.
- Clinical research enrolment criteria. Drug trials, dietary studies, and longitudinal cohorts routinely stratify or exclude by BMI.
A brief history of BMI
The "Quetelet index" was proposed by Adolphe Quetelet, a Belgian astronomer and social statistician, in 1832. Quetelet was not a physician and explicitly did not intend the index to diagnose individual health — his goal was to characterize the "average man" as a statistical tool for actuarial and census work. For more than a century the index lived in statistics departments and actuarial tables, unknown to clinical medicine.
In 1972, physiologist Ancel Keys's landmark paper "Indices of relative weight and obesity" reintroduced the Quetelet index into epidemiology under the new name Body Mass Index, demonstrating that it correlated better with skin-fold-measured body fat than several competing indices in the populations Keys studied (predominantly middle-aged American men). The WHO adopted BMI as the standard anthropometric obesity screen in 1995, and from there it diffused into every general-practice clinic, insurance form, and school health program on Earth.
Modern critique of BMI — that it was derived from a non-representative population, that it conflates muscle and fat, that its thresholds ignore ethnicity and age — has prompted many research groups to propose alternatives (ABSI, waist-to-height ratio, body-roundness index). None has displaced BMI, largely because BMI's advantages (free, universal, requires no equipment, interpretable by any clinician) still outweigh its acknowledged weaknesses for population-scale screening.
Measurements that should be reported alongside BMI
Because BMI's limitations are well-documented, clinical and research practice has converged on a small set of complementary measures that, together with BMI, give a much more reliable picture of body composition and cardiometabolic risk. If you're tracking your health seriously, don't rely on BMI alone.
- Waist circumference. Measured at the midpoint between the lowest rib and the iliac crest. WHO action levels: ≥94 cm (men) or ≥80 cm (women) indicates increased risk; ≥102 cm (men) or ≥88 cm (women) indicates substantially increased risk. Waist circumference captures central (visceral) adiposity, which is more cardiometabolically harmful than subcutaneous fat, and it distinguishes muscular-but-big from fat-and-big far more reliably than BMI.
- Waist-to-hip ratio (WHR). Waist circumference divided by hip circumference. WHO risk thresholds: >0.90 (men) or >0.85 (women) indicates increased abdominal adiposity. WHR is a particularly strong predictor of cardiovascular events in some populations — better than BMI alone in large prospective studies.
- Waist-to-height ratio (WHtR). A rising star in public health screening: waist ÷ height, with a universal cutoff of 0.5 ("keep your waist less than half your height") that works across ethnicities, sexes, and ages. Several studies find WHtR outperforms both BMI and WHR at predicting metabolic syndrome.
- Body fat percentage. Measured by skinfold caliper, bioelectrical impedance (BIA), air-displacement plethysmography (Bod Pod), or DEXA scan. Healthy ranges vary by age and sex. See our body-fat calculator.
- Blood-based metabolic markers. Fasting glucose, HbA1c, lipid panel (total cholesterol, LDL, HDL, triglycerides), blood pressure. These directly measure cardiometabolic state rather than inferring it from body shape.
A person with BMI 26, waist 82 cm, WHtR 0.45, body fat 18%, and normal lipids is in excellent health despite being "overweight" by BMI. A person with BMI 24, waist 95 cm, WHtR 0.56, body fat 32%, and elevated triglycerides is at substantially higher metabolic risk despite being "normal weight" by BMI. This is the phenomenon clinicians call TOFI — "Thin Outside, Fat Inside" — and it's precisely what BMI alone misses.
Sources & further reading
- World Health Organization — Obesity and overweight (official BMI classification and global prevalence data).
- US CDC — About Adult BMI (patient-facing explanations and calculator).
- NIH NHLBI — Calculate Your BMI (US National Institutes of Health reference).
- Wikipedia — Body Mass Index (history, alternative indices, population-specific thresholds).
- Keys, A., Fidanza, F., Karvonen, M.J., Kimura, N., Taylor, H.L. (1972). "Indices of relative weight and obesity." Journal of Chronic Diseases, 25(6), 329–343. The paper that introduced the name "Body Mass Index."
- WHO Expert Consultation (2004). "Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies." The Lancet, 363(9403), 100+–163.
Frequently Asked Questions
For adults, the WHO-defined normal range is 18.5 to 24.9 kg/m². Below 18.5 is underweight, 25 to 29.9 is overweight, and 30 or above is obese. These thresholds apply to most adult populations; WHO uses lower thresholds for Asian populations (overweight from 23).
BMI = weight (kg) ÷ height² (m²). In imperial units: BMI = 703 × weight (lb) ÷ height² (in²). Example: 70 kg at 1.75 m gives 70 ÷ (1.75 × 1.75) = 70 ÷ 3.0625 = 22.9, which is in the normal range.
No. BMI cannot distinguish muscle from fat, so highly muscular individuals (athletes, bodybuilders, rugby and NFL players) often land in the overweight or obese BMI category despite low body-fat percentage. For this group, body-fat percentage via skinfold, bioimpedance, or DEXA scan is far more informative.
BMI 25.0 to 29.9 is WHO-classified as overweight (pre-obese). BMI 30 or higher is obese, subdivided into class I (30–34.9), class II (35–39.9), and class III (≥40, also called morbid or severe obesity).
This calculator uses the adult formula and adult thresholds. For anyone under 20, use the age- and sex-adjusted BMI percentile charts from the CDC or WHO. A BMI that would be "normal" in an adult can be overweight in a child, and vice versa — fixed adult thresholds produce misleading results.
Every 3 to 6 months under consistent conditions (same time of day, same scale, same clothing state) is sufficient to detect meaningful change. More frequent checks reveal measurement noise rather than real trend. The 3-to-6-month window matches the timescale of meaningful body-composition change.
No. Expected pregnancy weight gain invalidates BMI thresholds. Use pre-pregnancy BMI only, and follow your obstetrician's gestational-weight-gain guidance (which itself depends on pre-pregnancy BMI).
Yes. At the same BMI, East and South Asian populations tend to have higher body-fat percentage and higher cardiometabolic risk than European populations. The WHO Western Pacific guidelines set the overweight threshold at 23 (not 25) for Asian adults. Clinical guidelines for South Asian patients often apply these lower thresholds.
Healthy-weight range = BMI 18.5 to 24.9 × height². For 1.75 m: min = 18.5 × 3.0625 = 56.7 kg, max = 24.9 × 3.0625 = 76.3 kg. The calculator displays this range automatically alongside your BMI result.
BMI is a reasonable screening measure at the population level. For individual health assessment, it should be combined with waist circumference or waist-to-hip ratio (for central adiposity), body-fat percentage, blood pressure, lipid panel, and fasting glucose. BMI alone cannot diagnose obesity or predict individual health outcomes.
Obesity class III (BMI ≥ 40), also called severe or morbid obesity, carries the highest BMI-associated health risk (cardiovascular disease, type 2 diabetes, obstructive sleep apnea, joint disease, some cancers). It is a criterion for bariatric surgery consultation in most health systems.
Yes, completely free. No signup, no account, no email required. Every calculation runs in your browser — your height and weight are never sent to our servers. The site is funded by non-intrusive display ads via Google AdSense; see our privacy policy.