Free BMI Calculator
Enter your height and weight to calculate your Body Mass Index instantly.
How It Works
BMI (Body Mass Index) is calculated as weight in kilograms divided by the square of height in meters: BMI = kg / m². Enter your measurements, press Calculate, and get an instant result with a category classification.
BMI Categories (WHO)
- Underweight · BMI below 18.5
- Normal weight · BMI 18.5 – 24.9
- Overweight · BMI 25 – 29.9
- Obese · BMI 30 and above
Frequently Asked Questions
Is BMI a perfect measure of health?
No. BMI is a simple screening tool based on height and weight. It does not account for muscle mass, bone density, age, sex, or body-fat distribution. Consult a healthcare professional for a complete health assessment.
Does it work for children?
This calculator uses the standard adult BMI formula. Child and teen BMI is age- and sex-specific and requires growth-chart percentile interpretation.
Is my data stored?
No. All calculations run entirely in your browser. Nothing is sent to any server.
The formula in two notations
Body Mass Index is defined as weight in kilograms divided by the square of height in metres: BMI = mass (kg) ÷ height (m)². The result is a number with units of kg/m², though everyday usage drops the unit and reports BMI as a bare decimal. A 70 kg adult who is 1.75 m tall has a BMI of 70 ÷ (1.75 × 1.75) ≈ 22.86.
When weight is in pounds and height in inches, the imperial form needs a conversion factor: BMI = 703 × mass (lb) ÷ height (in)². The 703 comes from the unit conversion (1 kg = 2.2046 lb, 1 m = 39.3701 in). A 154 lb adult who is 5'9\" (69 in) tall has BMI = 703 × 154 ÷ 69² ≈ 22.74. The two forms produce essentially the same answer; the small rounding difference is well below any clinically meaningful threshold.
A short history: Quetelet, Keys, and the long road from population statistic to medical metric
The formula was first published by the Belgian astronomer-statistician Adolphe Quetelet in 1832, in a paper on the average heights and weights of Belgian and French populations. Quetelet noticed that adult body weight scales roughly as the square of height, a better empirical fit than the cube (which would be expected if humans scaled isometrically). He used the ratio purely as a population descriptor, never as a measure of obesity or individual health, and his data came almost exclusively from 19th-century Western European subjects.
For 140 years the formula was used occasionally in academic anthropometry as the \"Quetelet Index.\" It did not enter clinical medicine. Insurance companies in the early 20th century relied on the Metropolitan Life height-and-weight tables (first published 1942, revised 1959 and 1983) to set premiums. The pivot came in July 1972 with a paper by Ancel Keys and colleagues in the Journal of Chronic Diseases, \"Indices of relative weight and obesity.\" Keys evaluated competing weight-for-height indices against skinfold-thickness measurements in 7,400 men from five countries; the Quetelet ratio came out as the best of a mediocre lot, and Keys proposed renaming it the \"body mass index.\"
Keys explicitly cautioned that the index was suitable for population studies, not individual diagnosis. That cautionary framing was almost immediately discarded. Through the 1970s and 1980s BMI was adopted by epidemiologists, then by life insurers, then by primary-care physicians as a screening tool. In 1985 a US NIH consensus panel adopted the first formal cutoffs (BMI > 27.8 for men, 27.3 for women). In 1998 the NIH harmonised with the WHO and lowered the obesity threshold to BMI ≥ 30 for both sexes while introducing the \"overweight\" category at 25-29.9. Roughly 29 million Americans changed weight categories overnight without gaining a pound.
The full WHO classification
| Category | BMI range (kg/m²) |
|---|---|
| Severe thinness | < 16.0 |
| Moderate thinness | 16.0 – 16.99 |
| Mild thinness | 17.0 – 18.49 |
| Normal range | 18.5 – 24.9 |
| Overweight (pre-obese) | 25.0 – 29.9 |
| Obese class I | 30.0 – 34.9 |
| Obese class II | 35.0 – 39.9 |
| Obese class III ("severe" obesity) | ≥ 40.0 |
The 18.5 / 25 / 30 thresholds are statistical conveniences, not biologically discrete boundaries. They were chosen because mortality risk in large epidemiological cohorts (e.g., the Lancet 2009 Prospective Studies Collaboration meta-analysis of 894,576 participants) shows a roughly J-shaped curve with the nadir around 22.5-25. The WHO acknowledges in its own technical reports that \"the cut-off points are necessarily arbitrary.\"
The 2004 Asian-population revision
In 2004 the WHO Expert Consultation acknowledged that the standard cutoffs underestimate cardiovascular and diabetes risk in many Asian populations, who tend to carry higher body-fat percentage and more visceral fat at lower BMIs than European-derived populations. Rather than universally lowering the thresholds, the consultation proposed additional public-health action points at BMI 23, 27.5, 32.5 and 37.5 for Asian populations. Singapore, Japan, China, India and South Korea use variants of these lower cutoffs in clinical practice. Since 2013 the NHS in the UK formally recommends the lower thresholds for adults of South Asian, Chinese, other Asian, Middle Eastern, Black African or African-Caribbean family origin.
Where BMI breaks down: the well-documented limits
Quetelet, Keys, the WHO, the CDC and the NHS all agree on this: BMI is a population-level screening proxy, not a diagnostic measure for an individual. The known failures:
- Doesn't account for muscle mass. Athletes and weightlifters routinely classify as \"overweight\" or \"obese\" because muscle is denser than fat. A rugby player with 12% body fat and a casual smoker with 28% body fat can share the same BMI.
- Doesn't differentiate fat distribution. Cardiometabolic risk depends heavily on where fat is. Visceral fat around abdominal organs is far more dangerous than subcutaneous fat on the hips and thighs. Two people with identical BMIs and very different waist measurements have very different risk profiles.
- Varies by ancestry. The original 1832 dataset was almost entirely white European. The 2004 WHO revision for Asian populations is one explicit acknowledgement; pediatric BMI percentile charts also stratify by population because growth curves differ.
- Doesn't apply to children. Child and teen BMI is age- and sex-specific and requires growth-chart percentile interpretation, not a fixed cutoff. Use a paediatric BMI-percentile calculator instead.
- Doesn't apply to pregnant women for diagnostic purposes. Pre-pregnancy BMI is the relevant baseline.
- Limited utility for older adults. Some research suggests slightly higher BMIs may be protective in adults over 65 (the so-called \"obesity paradox\"); the standard cutoffs are calibrated to mid-life adults.
- Doesn't apply to amputees or people with significantly atypical body proportions.
- Non-linear scaling. Quetelet noticed weight scales as the square of height, but tall and short people aren't perfectly proportional. The index slightly under-counts very tall people and over-counts very short ones.
Modern alternatives and complements
- Waist circumference, a direct measurement that correlates with visceral fat. WHO thresholds: above 102 cm (men) / 88 cm (women) for high cardiovascular risk in European-derived populations; lower in Asian-specific guidelines.
- Waist-to-height ratio, Margaret Ashwell's \"your waist should be less than half your height\" rule (waist ÷ height < 0.5). Simple, robust across ancestries, and more strongly correlated with cardiometabolic risk than BMI in most studies.
- Waist-to-hip ratio, a longer-standing measure that captures fat distribution. WHO thresholds 0.90 (men) / 0.85 (women).
- Body Roundness Index (BRI), a 2013 metric (Thomas et al.) that combines waist circumference and height into a single score. A 2024 JAMA Network Open paper found BRI predicted all-cause mortality more accurately than BMI in a large US cohort.
- Direct body-fat measurement, DEXA scan (clinical gold standard), bioelectrical impedance (consumer scales, less accurate), skinfold calipers (cheap, technique-dependent), or hydrostatic weighing (research only).
A note on critique and context
BMI has been the subject of substantive academic and public critique. Lindo Bacon's 2010 book Health at Every Size argued that weight-focused medicine is empirically weak and stigmatising, and proposed a behaviour-and-wellness focus instead. Sabrina Strings's 2019 book Fearing the Black Body: The Racial Origins of Fat Phobia traced the cultural origins of fatphobia in the West, including the role of 19th-century European data in shaping a metric now applied universally. The American Medical Association formally adopted a 2023 statement acknowledging BMI's limitations and recommending it be used \"in conjunction with other valid measures of risk such as, but not limited to, measurements of visceral fat, body adiposity index, body composition, relative fat mass, waist circumference and genetic / metabolic factors.\"
The takeaway: BMI is a quick, no-equipment screening number, useful as one signal among many. It is not a diagnosis, and an individual reading does not warrant alarm or complacency on its own. For real health decisions, talk to a healthcare professional who can interpret BMI alongside blood pressure, fasting glucose, lipid panel, family history, fitness, and lifestyle.
More questions
Why does the formula divide by height squared and not cubed?
Because Quetelet's empirical observation in 1832 was that weight scales as height-squared, not height-cubed (the cube is what you'd expect if humans grew isometrically, keeping the same shape as they got bigger). Real adult populations don't grow isometrically; we get proportionally narrower with height, so the squared denominator is a closer fit. The mathematician Nick Trefethen has argued that an exponent closer to 2.5 would be even better, but no major health body has adopted this.
My BMI is in the \"overweight\" category but I feel fine, should I be worried?
Not on the basis of BMI alone. As discussed above, the 25-29.9 cutoff is a population-level statistical convenience, and individual outcomes vary widely with muscle mass, fat distribution, ancestry, age and lifestyle. A more useful exercise: pair BMI with a waist measurement (Ashwell's \"less than half your height\" rule), a recent fasting glucose, and a blood-pressure reading. Bring all four to a primary-care visit.
Should I use the Asian-specific cutoffs?
If you're of South Asian, East Asian, Southeast Asian, Middle Eastern, Black African or African-Caribbean family origin, the lower thresholds (overweight at BMI ≥ 23, obese at ≥ 27.5) are the recommended interpretation by the NHS and many Asian national health authorities since around 2013. If you're of European ancestry, the standard 25 / 30 cutoffs apply. If you have mixed ancestry or are unsure, ask your healthcare provider.
Is BMI useful at all then?
Yes, as a free, no-equipment, fast screening number. Population-scale studies still find that BMI tracks all-cause mortality reasonably well at the cohort level, which is why it remains the default in epidemiology. The legitimate complaint isn't that BMI is meaningless; it's that BMI alone is too narrow a basis for individual medical decisions, and it has been historically applied as if it were more than a screening signal.
Does anything get sent to a server?
No. The formula is two multiplications and a division, computed in your browser. Your height, weight and the result never leave the page; the tool works offline once it's loaded.