Body Mass Index was invented in 1832 by a Belgian astronomer named Adolphe Quetelet who wanted a single number to describe the "average man." He had no clinical agenda — he was a statistician building population curves. Almost two centuries later, his weight-divided-by-height-squared has become the most-used and most-criticized health number in the world.[1]
BMI is on your medical chart, in your insurance underwriting, and inside the cutoffs the U.S. Preventive Services Task Force uses to decide whether your doctor should recommend behavioral weight-loss therapy.[2] About 73.6% of U.S. adults aged 20 and older are classified by BMI as overweight or obese, and roughly 40.3% meet the obesity cutoff.[3] Those numbers shape national policy, drug approvals, and surgical eligibility.
BMI is also, by the unanimous 2023 acknowledgement of the American Medical Association, an imperfect way to measure body fat that systematically misreads athletes, older adults, and people of non-European descent.[4] In January 2025 the Lancet Diabetes & Endocrinology Commission on Clinical Obesity went further and recommended that BMI no longer be used as a stand-alone diagnostic.[5]
This guide is the one we wish came pre-printed with every annual checkup. It covers the formula, the categories, the populations where BMI fails, the measurements clinicians now pair with it, three case studies, and an action checklist for what to do when your number falls in any category. When you are ready to run your own, the CalcLeap BMI calculator takes care of the arithmetic in metric or imperial.
⚖️Run your own BMI in 5 seconds
Free, instant, no signup — handles imperial and metric units.
What BMI actually is
BMI is a ratio. It takes your body mass and divides it by your height squared. That is it. There is no body-fat measurement, no muscle measurement, no fat-distribution measurement, and no age or sex adjustment in the formula itself. Two people with identical BMIs can be wildly different physiologically — a fact the formula was never designed to deny.
What BMI does well is rank a large population from leanest to heaviest in a way that, on average, correlates with body-fat percentage and with downstream health outcomes. The Centers for Disease Control and Prevention notes that BMI "is a screening tool, but it does not diagnose the body fatness or health of an individual."[6] That sentence is doing enormous work. BMI is to a clinical exam what a credit score is to a loan application: a fast number that earns a closer look, not the verdict itself.
Quetelet originally called the formula the "Quetelet Index." It was renamed Body Mass Index by physiologist Ancel Keys in a 1972 paper that compared several weight-height ratios and found Quetelet's to track skinfold-measured body fat better than the alternatives.[7] Keys recommended it for population studies. He was explicit that it was a research instrument, not a clinical one. The drift from epidemiology tool to clinic-room verdict happened over the following decades.
The formula (metric and imperial)
The metric form is the original and the cleaner one:
The imperial form needs a conversion constant because pounds and inches aren't on a base-10 footing with kilograms and meters:
The 703 comes from the algebra of converting pounds to kilograms (÷ 2.2046) and inches to meters (× 0.0254 then squared). It produces the same number as the metric formula, give or take a rounding artifact in the second decimal.
The intuition
BMI assumes weight should scale with the square of height. In reality, the human body is roughly three-dimensional, so a strict cube would be more accurate. That single mathematical choice is why BMI systematically misjudges very tall and very short people — tall people tend to score slightly higher than their actual body composition warrants; short people slightly lower. The error is small but real.
If you want to do it by hand: convert your height to meters (one inch = 0.0254 m, one foot = 0.3048 m), convert weight to kilograms (one pound = 0.4536 kg), then divide weight by the square of height. A 5'9" (1.7526 m) adult weighing 170 lb (77.1 kg) has a BMI of 77.1 ÷ (1.7526)2 = 77.1 ÷ 3.0716 = 25.1 — just barely in the "overweight" band by WHO cutoffs.
The WHO BMI categories — and the Asian-specific cutoffs
The World Health Organization adult BMI categories, adopted in 1995 and reaffirmed in the 2000 obesity consultation, are the global standard:[8]
| Category | BMI range (kg/m²) | Notes |
|---|---|---|
| Underweight | < 18.5 | Severe < 16.0; moderate 16.0–16.9; mild 17.0–18.4 |
| Normal weight | 18.5 – 24.9 | Reference range; lowest all-cause mortality on most curves |
| Overweight (pre-obese) | 25.0 – 29.9 | Risk of cardiovascular and metabolic disease rises |
| Obesity, class I | 30.0 – 34.9 | Substantially elevated disease risk |
| Obesity, class II | 35.0 – 39.9 | Severe risk; intensive intervention indicated |
| Obesity, class III (severe) | ≥ 40.0 | Bariatric surgery typically eligible |
Source: WHO Technical Report Series 894 and BMI classification page; thresholds apply to adults aged 20+.[8]
For populations of South Asian, East Asian, and Pacific Islander descent, the WHO Expert Consultation in 2004 noted that cardiometabolic risk — type 2 diabetes, hypertension, dyslipidemia — climbs at a lower BMI than in white European populations. India, China, Japan, and Singapore have all adopted lower cutoffs as a result. Common Asian-specific thresholds use 23.0 as the overweight boundary and 27.5 as the obesity boundary.[9] The U.S. CDC and most American clinical software still use the standard cutoffs, which means a South Asian patient with a "normal" U.S. BMI of 24 may already carry cardiometabolic risk equivalent to a white patient at BMI 27.
How to read your own BMI — worked examples
Here are five worked examples spanning the categories. All use the metric formula; the imperial conversions are shown for cross-reference.
| Profile | Height | Weight | Calculation | BMI | Category |
|---|---|---|---|---|---|
| Young woman | 5'4" / 1.63 m | 105 lb / 47.6 kg | 47.6 ÷ 2.656 | 17.9 | Underweight |
| Average U.S. man | 5'9" / 1.75 m | 200 lb / 90.7 kg | 90.7 ÷ 3.063 | 29.6 | Overweight |
| Average U.S. woman | 5'4" / 1.63 m | 171 lb / 77.6 kg | 77.6 ÷ 2.656 | 29.2 | Overweight |
| NFL running back | 6'0" / 1.83 m | 220 lb / 99.8 kg | 99.8 ÷ 3.349 | 29.8 | Overweight (misleading — see below) |
| Sedentary 70-year-old | 5'7" / 1.70 m | 148 lb / 67.1 kg | 67.1 ÷ 2.890 | 23.2 | Normal (potentially misleading) |
Average U.S. adult weights from CDC NHANES anthropometric reference data; running-back proportions are illustrative.[10]
The average American adult sits in the "overweight" band by BMI. That is not an editorial choice — it is the arithmetic of a country whose average adult man weighs 200 pounds and whose average adult woman weighs 171, with median heights near 5'9" and 5'4" respectively.[10] A "normal" BMI is by definition a minority status in the contemporary United States.
🍎Want the calorie number too?
Our calorie calculator estimates your maintenance, deficit, and surplus targets using Mifflin–St Jeor.
Where BMI breaks down: the four predictable failure modes
BMI's failures are not random. They cluster in four populations, and they are well-documented.
1. Muscular athletes
Muscle is about 18% denser than fat. A 6'0" athlete carrying 220 pounds of bodybuilder-grade musculature lands at BMI 29.8 — the upper edge of "overweight" — despite a body-fat percentage that may be under 12%. Studies of professional rugby and football players consistently find that more than half are classified as overweight or obese by BMI alone, with healthy or low body-fat measurements.[11] The formula treats every kilogram the same, but a kilogram of contractile tissue is metabolically and prognostically nothing like a kilogram of visceral adipose.
2. Older adults with sarcopenia
The opposite failure happens in aging. Between roughly 30 and 80, an adult loses an average of 0.5%–1% of muscle mass per year, a process called sarcopenia, while body fat tends to rise. An older adult may hold a stable weight (and therefore a stable BMI) while their body composition shifts dramatically toward fat. A "normal" BMI in a 70-year-old can mask sarcopenic obesity, which carries higher cardiometabolic and frailty risk than the same BMI in a 30-year-old.[12]
3. Different ancestry, different risk
BMI was derived from 19th-century white European populations. The risk-vs-BMI curve is not the same for everyone. South and East Asian populations develop type 2 diabetes and hypertension at lower BMIs because of a tendency toward higher visceral fat at any given weight. Many Black populations, by contrast, carry higher lean mass and bone density at any given BMI, meaning the same number reflects somewhat less fat. The WHO's 2004 expert consultation acknowledged that population-specific cutoffs are appropriate for clinical use.[9]
4. Children and teens
Children's body composition changes continuously with growth. A BMI of 22 in a six-year-old is alarming; the same BMI in a 16-year-old can be perfectly normal. For ages 2 to 19, BMI must be converted to a percentile against CDC sex-and-age growth charts. The CDC defines overweight as the 85th–94th percentile and obesity as the 95th percentile or above.[13] A child's calculated BMI number on its own is uninterpretable. Use the CDC BMI-for-age calculator, or a pediatrician's printed growth chart, never the adult cutoffs.
The four signals to ignore your BMI alone
You're an athlete with visible musculature; you're over 65 and your weight has been stable for years; you have South, East, or Southeast Asian ancestry; you are under 20. In any of those situations, your BMI number is at best an opening data point — not a verdict.
Three case studies that show the gap
Case 1: The "overweight" defensive lineman
Marcus is 6'2", weighs 245 pounds, and plays Division I football. His BMI is 31.4 — solidly in WHO obesity class I. A DEXA body-composition scan puts him at 11% body fat, with lean mass in the 96th percentile for his age and height. His fasting glucose, A1C, lipid panel, and blood pressure are all in the optimal range. By BMI he is a candidate for intensive behavioral weight-loss intervention; by every other measurement he is one of the metabolically healthiest 25-year-olds his physician has tested. He is the textbook example of why BMI must not be used in isolation.
Case 2: The "normal-weight" retiree
Susan is 67, 5'4", and weighs 145 pounds — a BMI of 24.9, right at the upper edge of the WHO normal range. She has not weighed herself in two years and considers herself "fine on weight." A body-composition scan shows 38% body fat and lean mass in the 12th percentile for her age. Her waist circumference is 34 inches. She has prediabetes (A1C 6.0%) and elevated triglycerides. Her BMI tells none of this story. Her waist measurement and lab panel tell all of it.
Case 3: The South Asian software engineer
Raj is 32, 5'8", and weighs 165 pounds — a BMI of 25.1, barely into "overweight" by U.S. cutoffs. By the Asian-specific cutoffs used in India and Singapore, he is in obesity class I. His waist circumference is 38 inches; his fasting glucose is 108 mg/dL (prediabetes range), and his triglycerides are 195. He has a strong family history of type 2 diabetes. A U.S.-style BMI reading dismisses him as borderline. A population-appropriate reading flags him for early intervention. The difference between those two readings is, in his case, the difference between catching insulin resistance at 32 versus diagnosing type 2 diabetes at 42.
The measurements that actually beat BMI alone
None of these replace BMI — they pair with it. Together they give a far better read on cardiometabolic risk than any single number.
| Measurement | What it captures | How to do it | Risk threshold |
|---|---|---|---|
| Waist circumference | Visceral (organ-surrounding) fat — the most dangerous kind | Tape measure at the level of the iliac crest, after a normal exhale | > 40 in (102 cm) men, > 35 in (88 cm) women per NHLBI[14] |
| Waist-to-hip ratio | Fat distribution (apple vs pear) | Waist circumference ÷ hip circumference | > 0.90 men, > 0.85 women per WHO |
| Waist-to-height ratio | Central adiposity, height-adjusted | Waist (in or cm) ÷ height (same units) | > 0.50 — the simple "keep your waist under half your height" rule |
| Body Roundness Index (BRI) | Volumetric body shape combining height and waist | Equation; available in clinical calculators | 2024 NHANES analysis: higher BRI predicted all-cause mortality better than BMI alone[15] |
| Body fat % (DEXA, BIA, skinfold) | Direct fat measurement | DEXA scan in a clinic; bioimpedance scale; calipers | > 25% men, > 32% women (ACE guidelines) typically flag elevated risk |
| Metabolic panel | Insulin resistance, dyslipidemia | Fasting blood draw at primary care visit | A1C ≥ 5.7% prediabetic; LDL ≥ 130, triglycerides ≥ 150 elevated |
The simplest single number to add to your BMI is your waist measurement. A 2020 IDF/AHA review found that adding waist circumference to BMI substantially improved prediction of cardiovascular events and type 2 diabetes versus BMI alone, especially in the BMI 25–30 range where the marginal information matters most.[16]
If you want one more, get the metabolic panel. A "normal" BMI with an A1C of 6.2 is a far more urgent clinical situation than an "obese" BMI with optimal labs. The lab panel sees the inside; BMI only sees the outside.
BMI in 2026: how clinical practice has shifted
Two formal statements have changed the standing of BMI in the last three years.
June 2023, American Medical Association. The AMA House of Delegates adopted new policy explicitly recognizing that BMI is "an imperfect way to measure body fat in multiple groups given that it does not account for differences across race/ethnic groups, sexes, genders, and age-span." The policy directed physicians to use BMI in conjunction with other valid measures of risk, such as visceral fat, body-adiposity index, body-fat measurements, and waist circumference, alongside genetic and metabolic factors.[4]
January 2025, Lancet Diabetes & Endocrinology Commission on Clinical Obesity. A 58-member international commission, chaired by Francesco Rubino of King's College London, recommended replacing BMI-alone obesity diagnoses with a two-tier framework: preclinical obesity (excess adiposity confirmed by BMI plus at least one anthropometric measure such as waist circumference, waist-to-hip ratio, or waist-to-height ratio, or by direct fat measurement — but without organ or functional impairment), and clinical obesity (the same body-composition criteria plus evidence of obesity-related organ dysfunction or daily functional limitation). The commission's framework has been endorsed by more than 75 medical organizations worldwide.[5]
The practical effect: BMI remains the entry point because it is fast, free, and reasonably correlated with body fat at the population level. But "you have a BMI of 32" is no longer a sufficient diagnosis. Your clinician should be confirming with at least one other anthropometric measure and looking for actual organ or functional involvement before assigning a clinical diagnosis of obesity.
What this changes for you
If your annual physical ends with "your BMI is X, here is your category" and nothing else, ask for a waist measurement and a metabolic panel. The current clinical standard, by both the AMA and the Lancet Commission, requires both before any obesity diagnosis is made.
An action checklist for this week
- Calculate your BMI now with the CalcLeap BMI calculator. Use accurate height (measured in the morning, no shoes) and accurate weight (morning, after restroom, no clothes). Note the category.
- Measure your waist. Tape measure at the level of the top of your hip bones, after a normal exhale, no sucking in. Above 40 inches (men) or 35 inches (women) flags elevated cardiometabolic risk regardless of BMI.
- Calculate your waist-to-height ratio. Waist ÷ height in the same units. Aim to keep it under 0.5. This single number captures more cardiometabolic risk information than BMI in most populations.
- Adjust for your context. If you are South or East Asian, mentally apply the 23/27.5 cutoffs. If you are over 65, consider that "normal" BMI can mask sarcopenic obesity — your waist matters more than your scale weight. If you are an active athlete, expect BMI to overstate; weigh body composition instead.
- Book a metabolic panel at your next primary care visit if you haven't had one in two years. A1C, fasting glucose, lipid panel, and blood pressure are the four numbers that actually drive 10-year cardiovascular risk.
- If your BMI is 30+ and labs are off, ask your clinician about the USPSTF-recommended intensive multicomponent behavioral interventions — they show consistent, modest, durable weight-loss benefits with minimal harm.[2] Pharmacotherapy and surgical options exist for those who meet criteria.
- Use the right tool for kids. Children and teens go through the CDC BMI-for-age percentile chart, never the adult cutoffs. Your pediatrician's office handles this automatically; you should not interpret a child's raw BMI number.
- Run the related numbers. Plug into our calorie calculator for a daily maintenance target, the body-fat calculator for a Navy-method estimate, and the ideal weight calculator for the Devine/Robinson healthy-range estimates.
Frequently asked questions
What is BMI in simple terms?
BMI (Body Mass Index) is your weight in kilograms divided by the square of your height in meters. It is a screening number that puts adults into one of four categories: under 18.5 underweight, 18.5–24.9 normal, 25.0–29.9 overweight, and 30.0 or above obese. It is not a diagnosis on its own — clinicians use it as a starting point, not an ending point.
What is the BMI formula?
Metric: BMI = weight (kg) / height (m)2. Imperial: BMI = 703 × weight (lb) / height (in)2. The 703 conversion constant accounts for the unit change and produces the same number to within rounding.
What is a healthy BMI for adults?
The WHO and CDC define the normal-weight range as a BMI of 18.5 to 24.9 for adults aged 20 and older. For people of South or East Asian descent, several guidelines suggest a lower overweight cutoff of 23.0 because cardiometabolic risk rises at a lower BMI in those populations.
Where does BMI fail?
BMI does not distinguish muscle from fat, does not measure where fat is stored, and was originally derived from white European populations. It systematically misclassifies muscular athletes as overweight, can miss high body-fat in older adults whose muscle has shrunk, and reads differently across racial and ethnic groups. In 2023 the American Medical Association formally acknowledged these limits and advised pairing BMI with other measures.
Is BMI accurate for children?
Not the way it is for adults. For ages 2–19, BMI must be converted to a percentile against same-age, same-sex CDC growth charts. A child at the 85th–94th percentile is overweight; 95th and above is obese. Use the CDC BMI-for-age calculator or a pediatrician's chart, never the adult cutoffs.
What is a better measurement than BMI?
There is no single replacement, but waist circumference, waist-to-hip ratio, and waist-to-height ratio all capture central (visceral) fat — the kind most linked to cardiovascular and metabolic disease. Body Roundness Index (BRI) combines height and waist into a single score that, in 2024 NHANES analyses, predicted mortality better than BMI alone. Body fat percentage from DEXA, skinfold, or bioimpedance is the most direct measurement when available.
Has BMI been replaced in clinical practice?
Not replaced — recontextualized. The January 2025 Lancet Diabetes & Endocrinology Commission on Clinical Obesity recommended diagnosing obesity using a combination of BMI plus at least one other anthropometric measure (waist circumference, waist-to-hip ratio, or waist-to-height ratio) or direct fat measurement, plus evidence of organ or functional impairment. BMI alone is now considered a screening tool, not a diagnostic one.
What should I do if my BMI is in the overweight or obese range?
Treat it as a signal to look closer, not a verdict. Measure your waist (a waist over 35 inches in women or 40 in men flags higher cardiometabolic risk per the NHLBI). Ask your primary care clinician for a metabolic panel — fasting glucose or A1C, lipid panel, blood pressure. The USPSTF recommends offering intensive, multicomponent behavioral interventions to adults with a BMI of 30 or higher. Pharmacotherapy and surgery are options for those who meet clinical criteria.
Methodology & sources
All calculations in this article use the standard BMI formula BMI = weight (kg) / height (m)2, with the imperial form BMI = 703 × weight (lb) / height (in)2. Adult category cutoffs follow the WHO classification (1995, reaffirmed 2000). Asian-specific cutoffs (23/27.5) follow the 2004 WHO Expert Consultation. Pediatric cutoffs follow CDC BMI-for-age percentile guidance. Prevalence figures use the most recent CDC NHANES release covering August 2021 through August 2023. Clinical-practice recommendations cite the AMA 2023 policy and the Lancet Diabetes & Endocrinology Commission on Clinical Obesity published January 2025. Worked examples were computed in Python and cross-checked by hand.
Sources cited:
- Eknoyan G. (2008). Adolphe Quetelet (1796–1874) — the average man and indices of obesity. Nephrology Dialysis Transplantation. academic.oup.com/ndt
- U.S. Preventive Services Task Force, Behavioral Weight Loss Interventions to Prevent Obesity-Related Morbidity and Mortality in Adults: Recommendation Statement (2018, reaffirmed). uspreventiveservicestaskforce.org
- CDC National Center for Health Statistics, Obesity and Severe Obesity Prevalence in Adults: United States, August 2021–August 2023, NCHS Data Brief No. 508. cdc.gov/nchs
- American Medical Association, AMA adopts new policy clarifying role of BMI as a measure in medicine (June 14, 2023). ama-assn.org
- Rubino F. et al. (2025). Definition and diagnostic criteria of clinical obesity. Lancet Diabetes & Endocrinology. Commission report, January 14, 2025. thelancet.com
- U.S. Centers for Disease Control and Prevention, About Adult BMI. cdc.gov/bmi
- Keys A. et al. (1972). Indices of relative weight and obesity. Journal of Chronic Diseases 25(6–7):329–343.
- World Health Organization, BMI Classification (Global Database on Body Mass Index). who.int
- WHO Expert Consultation (2004). Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. The Lancet 363:157–163. thelancet.com
- CDC NCHS, Anthropometric Reference Data for Children and Adults: United States, 2015–2018, Vital and Health Statistics Series 3, No. 46. cdc.gov/nchs
- Provencher M.T. et al. (2018). The impact of body mass index on the prevalence and severity of injury and concussion in National Collegiate Athletic Association Division I football players. American Journal of Sports Medicine. PubMed indexing on NIH.gov: pubmed.ncbi.nlm.nih.gov
- Batsis J.A., Villareal D.T. (2018). Sarcopenic obesity in older adults: aetiology, epidemiology and treatment strategies. Nature Reviews Endocrinology. NIH PMC: ncbi.nlm.nih.gov/pmc
- U.S. Centers for Disease Control and Prevention, About Child & Teen BMI. cdc.gov/bmi child & teen
- National Heart, Lung, and Blood Institute (NHLBI), Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults — waist circumference cutoffs. nhlbi.nih.gov
- Zhang X. et al. (2024). Body Roundness Index and all-cause mortality among U.S. adults. JAMA Network Open 7(6):e2415051. jamanetwork.com
- Ross R. et al. (2020). Waist circumference as a vital sign in clinical practice: a Consensus Statement from the IAS and ICCR Working Group on Visceral Obesity. Nature Reviews Endocrinology 16:177–189. nature.com
This article is educational. It is not a substitute for personalized medical advice. Always consult a licensed clinician for diagnosis and treatment of any health condition. Read our editorial process →