BMI Formula Explained: Calculation, Categories, History
Core Conclusion
BMI = weight divided by height squared (metric). The World Health Organization and US Centers for Disease Control and Prevention classify BMI values as follows: below 18.5 underweight, 18.5 through 24.9 normal, 25.0 through 29.9 overweight, and 30.0 and above obese. The index was created in 1832 by Adolphe Quetelet and formally adopted for broad population-level use in 1972 by Ancel Keys.
Body Mass Index appears on nearly every public health dataset in circulation today. Hospitals record it on intake forms, epidemiologists use it to describe populations, and statistical agencies publish its distribution in regular reports. Despite this ubiquity, the origins, mathematics, and classification boundaries of BMI remain poorly understood by many readers. This page documents the formula from first principles, walks through its 190-year intellectual history, catalogs all major official classification standards in current use worldwide, and summarizes the known limitations as stated in peer-reviewed literature and government public health guidance.
Readers interested in computing their own BMI value can use the [Adult BMI Calculator + ../tools/bmi-calculator.html] which implements the exact formulas described below. For younger populations, the [Teen BMI Calculator + ../tools/bmi-teen-calculator.html] applies CDC pediatric percentile growth curves. Height-to-weight reference ranges derived from BMI intervals are available in the [Standard Weight Chart + ../tools/standard-weight-chart.html].
The Quetelet Index Origin
The mathematical construct that would later be renamed BMI first appeared in the work of Lambert Adolphe Jacques Quetelet, a Belgian mathematician, astronomer, and statistician active in the first half of the nineteenth century. Quetelet described the weight-to-height-squared relationship in his 1835 treatise "Lettre sur l'homme et le développement de ses facultés, ou Essai de physique sociale" (translated: A Treatise on Man and the Development of His Faculties, or Essay on Social Physics).
Quetelet's project was explicitly statistical rather than medical. His broader work sought to define the characteristics of what he termed "l'homme moyen" — the average man — as a basis for describing human population distributions. The weight-to-height-squared ratio was one among many anthropometric relationships he documented in cross-sectional population data. Crucially, Quetelet did not propose the ratio as a diagnostic or clinical tool for any individual[WHO].
The index remained primarily a statistical curiosity for well over a century. Its modern trajectory began in the early 1970s when the American physiologist Ancel Keys and his colleagues published a 1972 paper titled "Indices of relative weight and obesity" in the Journal of Chronic Diseases. Keys reviewed several competing weight-for-height indices against observed body fat measurements and concluded that the Quetelet index correlated most closely with adiposity across a range of adult populations. In this same paper, Keys formally proposed the name "Body Mass Index" or BMI[Keys 1972].
In 1985, the US National Institutes of Health Consensus Development Conference on Obesity adopted BMI as a standard population metric. The World Health Organization followed in 1995 with the publication of its first formal classification scheme, and expanded on this work with the 2000 WHO Global Database on BMI report[WHO 1995, WHO 2000]. The Centers for Disease Control and Prevention in the United States aligns with WHO ranges for adult populations 20 years of age and older[CDC].
Formula Mathematics
BMI is a unit-density ratio expressed in units of kilograms per meter squared. Two equivalent computational forms exist, one for the metric system (kilograms and meters) and one for the imperial system (pounds and inches). The imperial form incorporates a unit-conversion constant of 703, which converts pound-per-inch-squared values into kilogram-per-meter-squared equivalents.
Metric Formula (SI Units)
BMI = weightkg / (heightm × heightm)
Where weight is measured in kilograms and height is measured in meters. The height value is squared by multiplying the height by itself.
Imperial Formula (US Customary Units)
BMI = (weightlb × 703) / (heightin × heightin)
Where weight is measured in pounds and height is measured in inches. The factor of 703 is derived from the following unit conversion: 1 kilogram equals 2.20462 pounds, and 1 meter equals 39.3701 inches. Squaring the meter-to-inch ratio and dividing by the pound-to-kilogram ratio yields 39.3701² / 2.20462 ≈ 703.07, which rounds to 703 for practical computation.
Worked Example 1 — Metric: 70 kg, 1.75 m
A 32-year-old adult with a body weight of 70 kilograms and a height of 1.75 meters.
BMI = 70 / (1.75 × 1.75) = 70 / 3.0625 = 22.86
Result: 22.86, which falls within the WHO normal category range of 18.5 through 24.9.
Worked Example 2 — Imperial: 180 lb, 5 ft 10 in (70 in)
A 45-year-old adult weighing 180 pounds with a height of 5 feet 10 inches (70 total inches).
BMI = (180 × 703) / (70 × 70) = 126,540 / 4,900 = 25.82
Result: 25.82, which falls within the WHO overweight category range of 25.0 through 29.9.
Worked Example 3 — Metric: 95 kg, 1.82 m
A 55-year-old adult weighing 95 kilograms with a height of 1.82 meters.
BMI = 95 / (1.82 × 1.82) = 95 / 3.3124 = 28.68
Result: 28.68, which falls within the WHO overweight category range.
International Classification Standards
Multiple official public health bodies publish BMI classification tables. The WHO global standard is the most widely referenced internationally. Several national authorities, particularly in East and Southeast Asian jurisdictions, publish region-specific thresholds reflecting local epidemiological data. All thresholds presented below are official standards issued by named government or intergovernmental public health agencies.
WHO 1995 / WHO 2000 Global Database on BMI Classification
The 1995 WHO "Obesity: Preventing and managing the global epidemic" report and the expanded 2000 WHO Global Database on BMI define the following adult classification for populations aged 20 years and older. This scheme is used by the United Nations system, most European ministries of health, and the US CDC for adult populations.
| WHO Category | BMI Range (kg/m²) |
|---|---|
| Underweight (Severe thinness) | Below 16.0 |
| Underweight (Moderate thinness) | 16.0 – 16.9 |
| Underweight (Mild thinness) | 17.0 – 18.4 |
| 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 | 40.0 and above |
Regional Public Health BMI Cutoff Standards
The following thresholds represent official standards published by named national health authorities. They are cataloged here for reference purposes, presented as factual records of what each issuing agency has publicly adopted.
| Issuing Authority | Overweight Threshold | Obesity Threshold |
|---|---|---|
| WHO Global Standard | 25.0 | 30.0 |
| China National Health Commission | 24.0 | 28.0 |
| Japan MHLW / JASSO | 25.0 | 25.0 |
| Korea CDC (KDCA) | 23.0 | 25.0 |
| Singapore MOH / HPB | 23.0 | 27.5 |
Known Limitations
BMI does not measure body fat, or adiposity, directly. This limitation is stated explicitly on the CDC BMI public information pages and in multiple peer-reviewed publications. BMI is a weight-to-height ratio that treats all kilograms identically, without distinguishing between lean body mass (muscle, bone, organ tissue) and adipose (fat) tissue.
Several published analyses of NHANES (National Health and Nutrition Examination Survey) data document patterns where two individuals with identical BMI values can carry substantially different proportions of body fat. Individuals with high lean mass, for example, may register the same BMI number as individuals with higher adipose mass but lower lean mass. NHANES-based studies also document age and sex interactions: for a given BMI value, older adults tend to carry more body fat than younger adults, and women tend to carry more body fat than men at the same BMI[NHANES, CDC].
BMI also contains no information about the anatomical distribution of adipose tissue within an individual. Epidemiological research consistently finds that adipose distribution patterns (central versus peripheral) correlate with population-level health outcomes independently of total BMI value. The BMI formula contains no waist-circumference or waist-to-hip term and therefore cannot capture this dimension of body composition.
These limitations are inherent to the metric's design. BMI was selected by Keys and subsequent adoption bodies specifically as a population screening tool — a single inexpensive number that requires only a scale and a stadiometer to compute. It was never designed as an individual diagnostic metric. The limitations section of any BMI discussion is therefore not a "flaw" discovered later by critics but a known property of the metric that was present from the moment of its formal adoption.
Adult vs Pediatric BMI Differences
The arithmetic BMI formula — weight divided by height squared — is identical for adults and children. The difference between adult and pediatric BMI lies entirely in how the resulting number is interpreted, not in how it is computed.
For adults aged 20 years and older, fixed category thresholds are applied uniformly. A BMI of 22.0 falls in the normal range for a 22-year-old adult, and it falls in the normal range for a 72-year-old adult, regardless of sex.
For individuals aged 2 through 19 years, the computed BMI number is compared against age- and sex-specific growth chart percentiles maintained by the CDC. The CDC 2000 BMI-for-age growth charts were constructed from longitudinal and cross-sectional anthropometric surveys of US children. Category ranges are defined in terms of percentiles rather than fixed BMI values: below the 5th percentile corresponds to the underweight range, the 5th through 84th percentile to the healthy weight range, the 85th through 94th percentile to the overweight range, and the 95th percentile and above to the obesity range[CDC Pediatric].
The percentile-based approach accounts for the natural changes in body composition that occur during human growth and development. Young children naturally carry higher adiposity levels that decline through childhood, then rise again through adolescence — trajectories that would be mischaracterized by fixed adult thresholds applied mechanically.
Why Population Screening and Not Individual Diagnosis
The CDC website on BMI states the following verbatim: "BMI is a screening tool. It does not diagnose body fatness or health. A trained health care provider performs appropriate assessments to evaluate an individual's health status and risks."[CDC BMI Quote]
This distinction between a screening tool and a diagnostic tool is central to the public health use of BMI. Screening tools are designed to be inexpensive, quick, and scalable across large populations, with the understanding that a subset of screening results will require follow-up with more precise and resource-intensive assessment methods. Diagnostic tools, by contrast, are designed to produce individual-level determinations directly.
In the specific context of BMI, population-level screening allows public health researchers and epidemiologists to describe trends across large groups — for example, the distribution of BMI values in a national population compared across decades, or differences between demographic subgroups. For any single individual, a BMI number is one data point among many that a qualified clinician considers alongside clinical examination, medical history, additional measurements, and other relevant information.