BMI calculator

The sum

BMI is weight in kg ÷ (height in m)². In pounds and inches it's weight ÷ height² × 703. Keys named it the body mass index in 1972, after comparing simple weight-for-height measures in 7,424 men.

Sources 3, 4, 19

The ranges

For most adults, NICE and the World Health Organization (WHO) agree: 18.5 to 24.9 is a healthy weight, 25 to 29.9 overweight, and 30 or more obesity, in classes from 35 and 40. Under 18.5 is underweight. WHO's ranges are the same for women and men and don't change with age.

Sources 1, 2

Lower thresholds for some backgrounds

For people of South Asian, Chinese, other Asian, Middle Eastern, Black African or African-Caribbean background, NICE and the NHS treat 23 to 27.4 as overweight and 27.5 or more as obesity, because health risks start at a lower BMI. In 1.47 million adults in England, the BMI with the same diabetes risk as 30 in White adults ranged from 23.9 to 28.1.

Sources 1, 6, 7

BMI isn't body fat

Across 25 studies, the usual BMI cut-offs missed half the people with excess body fat. It can't tell muscle from fat, so muscular people can score high, and it can't show where you carry fat.

Sources 8, 9, 10, 5

What it does and doesn't predict

Across 239 studies, death rates were lowest at a BMI of 20 to 25 and rose below and above that. For one person, though, BMI adds little to heart risk once blood pressure, diabetes and cholesterol are known, and being fit cuts much of the risk linked to a higher BMI.

Sources 11, 12, 13

Who it isn't for

Not for under-18s, whose BMI is judged on growth centiles, or in pregnancy. No guideline uses different ranges for older adults, but NICE asks for caution from 65; separate low-BMI cut-offs for over-70s exist only for diagnosing malnutrition.

Sources 5, 14, 15, 16, 17

The formula

BMI = weight in kg ÷ (height in m)². WHO's own example is 70 kg and 1.75 m, which gives 22.9. In US units, the US Centers for Disease Control and Prevention (CDC) writes it as "weight (pounds)/[height (in)]2 x 703". The 703 is a rounding of 703.07 (our arithmetic), so the calculator converts feet, inches, stones and pounds exactly instead, and both kinds of units give the same answer.

Sources 3, 2, 4, 5

Units and rounding

Choose cm and kg, feet and stones, or feet and pounds. Heights from 120 to 230 cm (3 ft 11 in to 7 ft 6 in) and weights from 30 to 300 kg work (our choice). The result is shown to one decimal place, as the NHS shows it, and the range is read from the number you see, so a BMI of 24.96 shows as 25.0 and is in the overweight range.

Source 5

The ranges for most adultsStrong evidence

These are NICE's words (recommendation 1.9.10). NICE has no band below 18.5; WHO calls under 18.5 underweight. NICE adds: "Use clinical judgement when interpreting the healthy weight category because a person in this category may nevertheless have central adiposity."

• Underweight (WHO): under 18.5
• Healthy weight: BMI 18.5 kg/m2 to 24.9 kg/m2
• Overweight: BMI 25 kg/m2 to 29.9 kg/m2
• Obesity class 1: BMI 30 kg/m2 to 34.9 kg/m2
• Obesity class 2: BMI 35 kg/m2 to 39.9 kg/m2
• Obesity class 3: BMI 40 kg/m2 or more

Sources 1, 2

The lower thresholdsStrong evidence

NICE's recommendation 1.9.11: "People with a South Asian, Chinese, other Asian, Middle Eastern, Black African or African-Caribbean background are prone to central adiposity and their cardiometabolic risk occurs at lower BMI, so use lower BMI thresholds as a practical measure of overweight and obesity: overweight: BMI 23 kg/m2 to 27.4 kg/m2; obesity: BMI 27.5 kg/m2 or above. For people in these groups, obesity classes 2 and 3 are usually identified by reducing the thresholds highlighted in recommendation 1.9.10 by 2.5 kg/m2."

Choose that background and the calculator uses this table. The healthy weight band ends where overweight starts, and the obesity classes come from taking 2.5 off NICE's general thresholds (our arithmetic):

• Underweight: under 18.5
• Healthy weight: 18.5 to 22.9
• Overweight: 23 to 27.4
• Obesity class 1: 27.5 to 32.4
• Obesity class 2: 32.5 to 37.4
• Obesity class 3: 37.5 or more

Source 1

Worked examples (our arithmetic)

Each BMI is shown to one decimal place, and its range is read from that number:

• 70 kg and 175 cm: 70 ÷ 1.75² = 22.857, shown as 22.9, the healthy weight range.
• 70 kg and 170 cm (the NHS's own example): 24.221, shown as 24.2. That's a healthy weight on the general ranges, and overweight on NICE's lower thresholds.
• 11 st 0 lb and 5 ft 7 in (69.85 kg and 170.18 cm): 24.120, shown as 24.1. With 703 the sum gives 24.117, which shows the same.
• 76.44 kg and 175 cm: 24.96, shown as 25.0, so the overweight range.
• 95 kg and 180 cm: 29.321, shown as 29.3. That's overweight on the general ranges, and obesity class 1 on the lower thresholds.

Sources 2, 5, 1

What the result shows

Your BMI and its range, said about the number in NICE's words: "Your BMI is in the healthy weight range". Under it is the table of ranges for the background you chose, with yours marked, and a folded box, "What the bands mean for your height". That box shows the weights at a BMI of 18.5 and 24.9 for your height (22.9 on the lower thresholds) in your units: at 170 cm, 53.5 to 72.0 kg, or about 8 st 6 lb to 11 st 5 lb (our arithmetic). It never shows a weight to lose.

Every result reminds you that BMI can't tell muscle from fat, and gives the Beat eating disorder helpline. Under 18.5 it adds when to see a GP, at 65 and over NICE's caution, and below 12 or above 90 a note to check your entries. Under 18, or if you tick that you're pregnant, it stops and says where to go instead. The next step links to the waist-to-height ratio calculator.

Sources 5, 1, 22, 36

How BMI came aboutStrong evidence

Adolphe Quetelet, a Belgian mathematician and statistician, described the index in 1832, having found that in adults "the weight increases as the square of the height". In 1972 Keys compared simple measures of weight for height in 7,424 men in 12 groups in five countries. Weight ÷ height² was "slightly better" than weight ÷ height at tracking body fat without tracking height, and Keys named it the body mass index, preferring it for its simple sum and its "applicability to all populations at all times".

Sources 18, 19

The rangesStrong evidence

NICE and WHO use the same ranges for most adults: healthy weight 18.5 to 24.9, overweight 25 to 29.9 and obesity from 30, in three classes. WHO says its values are "age-independent and the same for both sexes", and the CDC uses the same numbers for adults "regardless of age, sex, or race".

Sources 1, 2, 3, 27

Lower ranges for some backgroundsStrong evidence

NICE uses 23 to 27.4 for overweight and 27.5 or more for obesity in people of South Asian, Chinese, other Asian, Middle Eastern, Black African or African-Caribbean background. In 1.47 million adults in England, the BMI with the same type 2 diabetes risk as 30 in White adults was 23.9 in South Asian, 26.6 in Arab, 26.9 in Chinese and 28.1 in Black adults.

Sources 1, 20, 7

BMI misses body fat in many peopleStrong evidence

In 25 studies of 31,968 people, the usual BMI cut-offs found only half of those with excess body fat (sensitivity 0.50, specificity 0.90). In 13,601 US adults, a BMI of 30 or more picked up 36% of men and 49% of women with high body fat. In 439 college students, a BMI of 25 or more labelled many men and female athletes overweight who weren't over-fat, while in women who didn't play sport it missed some with high body fat. NICE asks for caution with BMI in people with high muscle mass.

Sources 8, 9, 10, 1

What it tells you about healthStrong evidence

In 3.95 million people who had never smoked, death rates were lowest at a BMI of 20 to 25. Against 22.5 to 25, the risk was 51% higher at 15 to 18.5, 7% higher at 25 to 27.5, 45% higher at 30 to 35 and 2.76 times as high at 40 to 60. But in 221,934 people, adding BMI or waist size to blood pressure, diabetes and cholesterol didn't improve heart risk prediction, and in 20 studies fit people with overweight or obesity had no clearly higher mortality than fit people of a healthy weight. Low BMI and high body fat were each linked with higher mortality in 54,420 adults who had scans.

Sources 11, 12, 13, 26

Over 65Strong evidence

No guideline changes the ranges, but NICE asks for caution from 65, because health conditions, loss of muscle and the possibly protective effect of a slightly higher BMI matter more. In 197,940 adults aged 65 and over, mortality was 19% higher at a BMI of 20 to 20.9 and 12% higher at 21 to 21.9 than at 23 to 23.9, and began to rise above 33.

Sources 1, 20, 21

Not for childrenStrong evidence

Under 18, BMI is compared with children of the same age and sex on growth centiles, not adult ranges. NICE classes overweight from the 91st centile, clinical obesity from the 98th and severe obesity from the 99.6th, using the Royal College of Paediatrics and Child Health (RCPCH) charts.

Sources 14, 1, 15

What the NHS saysStrong evidence

The NHS: "If you are from a South Asian, Chinese, other Asian, Middle Eastern, Black African or African-Caribbean background you're at risk of health problems with a slightly lower BMI, so: a BMI of 23 to 27.4 is overweight; a BMI of 27.5 or higher is obese." Its adult calculator asks about your background "because people from an Asian, Black African, African-Caribbean or Middle Eastern ethnic background have a higher chance of developing health problems at a lower BMI".

Sources 6, 5

Why NICE lowered themStrong evidence

NICE's committee found that "studies in people of South Asian and Chinese family backgrounds showed an increased risk at a BMI of 21 kg/m2 to 26 kg/m2, whereas people from White family backgrounds showed increased risks at 25 kg/m2 to 29 kg/m2". For Middle Eastern (Arab and Iranian), Black African, Black Caribbean and other Asian (Japanese, Korean and Thai) family backgrounds, risk rose at values that "ranged from 21 kg/m2 to 30 kg/m2 but most were below 25 kg/m2". There was no direct evidence for obesity classes 2 and 3, so taking about 2.5 off was the committee's consensus. It also noted a lack of evidence for some ethnic groups, and few UK studies.

Source 20

What the research foundEmerging evidence

In 1.47 million adults in England, followed for a median of 6.5 years, the BMI with the same type 2 diabetes risk as 30 in White adults was 23.9 in South Asian, 26.6 in Arab, 26.9 in Chinese and 28.1 in Black adults. A WHO expert consultation in 2004 found that risk in Asian populations started anywhere from 22 to 25, and named extra action points at 23.0, 27.5, 32.5 and 37.5 while keeping WHO's international ranges. For the same body fat, age and sex, BMI was 1.9 to 4.6 units lower in Chinese, Ethiopian, Indonesian and Thai groups than in Caucasian groups.

Sources 7, 23, 24

If your background is mixed or not listed

Neither NICE nor the NHS gives separate advice for mixed backgrounds (our reading), so ask your GP which ranges suit you. The waist-to-height ratio uses the same bands for every ethnic background, which makes it a useful second check.

Source 1

Fat or muscleStrong evidence

BMI weighs everything together. The NHS: "if you have a lot of muscle, you may be classed as overweight or obese despite having low body fat". In 13,601 US adults, BMI couldn't tell body fat from lean mass between 25 and 29.9, and it did worse with age. NICE: "Interpret BMI with caution in adults with high muscle mass because it may be a less accurate measure of central adiposity in this group."

Sources 8, 9, 10, 5, 1

Women, men and ageStrong evidence

The ranges are the same for everyone, but body fat at a given BMI isn't. WHO: "the percentage of body fat mass increases with age up to 60-65 years in both sexes, and is higher in women than in men of equivalent BMI". So the same BMI can mean more body fat for an older woman than for a young man.

Sources 2, 27

Where you carry fatStrong evidence

NICE says to use BMI "with caution because it is not a direct measure of central adiposity", the fat around your middle, and pairs it with the waist-to-height ratio below a BMI of 35. In UK survey data on 1,453 adults, people with a healthy BMI but a ratio of 0.5 or more had worse blood fats, blood sugar and blood pressure than those with a healthy BMI and a smaller waist.

Sources 1, 25

FitnessEmerging evidence

In 20 studies of 398,716 observations, unfit people had 2 to 3 times the risk of dying of fit people of a healthy weight, at any BMI, while fit people with overweight or obesity had no clearly higher risk. These studies show links, not proof, but they're a good reason to train for fitness whatever your BMI.

Sources 13, 12

What NICE saysStrong evidence

"Interpret BMI with caution in people aged 65 and over, taking into account comorbidities, conditions that may affect functional capacity and the possible protective effect of having a slightly higher BMI when older." NICE's committee explains that functional capacity may be reduced by "conditions such as age-related spinal disorders or sarcopenia" (loss of muscle), and that a slightly higher BMI can be protective because older people with one are "less likely to be experiencing undernutrition". Enter an age of 65 or over and the calculator adds this caution.

Sources 1, 20

No different rangesStrong evidence

WHO's ranges are "age-independent", and the CDC uses them "regardless of age". Some calculators offer separate ranges for older adults, but we found no guideline that sets them. The one age-specific cut-off we found is for diagnosing malnutrition: the Global Leadership Initiative on Malnutrition (GLIM) criteria count a BMI under 20 below age 70, or under 22 from 70, as one sign (under 18.5 and 20 in Asia), and need a second sign as well.

Sources 2, 27, 17

What studies of older adults foundEmerging evidence

In 32 studies of 197,940 adults aged 65 and over, mortality was 12 to 19% higher at a BMI of 20 to 21.9 than at 23 to 23.9, and began to rise above 33. In a larger analysis, the extra risk for each 5 points of BMI above 25 was smaller in people measured at 70 to 89 than at 35 to 49. These are studies of groups, not ranges for you.

Sources 21, 11

If you're losing weight without trying

The NHS says to speak to a GP if you've lost weight suddenly, if you've lost weight without changing your diet or exercise routine, or if you use the control of food to cope with feelings and other situations.

Source 22

Use a calculator built for childrenStrong evidence

The NHS: "It's important to use the right calculator for adults (aged 18 and over) or children and teenagers (aged between 2 and 17)." Its child calculator needs a date of birth, sex at birth and the date of measuring, and "uses the same data sources as the RCPCH (Royal College of Paediatrics and Child Health) Digital Growth Charts". Enter an age under 18 here and the calculator stops and links to it.

Sources 28, 14

How children's BMI is judged in the UKStrong evidence

NICE asks for charts that are "appropriate for children and young people" and "adjusted for age and sex", and classes overweight from the 91st centile, clinical obesity from the 98th and severe obesity from the 99.6th. The RCPCH chart adds that a BMI below the 2nd centile "is unusual and may reflect undernutrition", though it can also just reflect a small build. The UK curves come from a large representative sample of British children.

Sources 1, 15, 29

Why adult ranges start at 18Strong evidence

International cut-offs for children were drawn from growth surveys in six countries so that "at age 18 years" they "passed through the widely used cut off points of 25 and 30". Before then, a child's BMI only makes sense against others of the same age and sex.

Source 30

In the USStrong evidence

The CDC uses percentiles for ages 2 through 19: underweight under the 5th, a healthy weight from the 5th to under the 85th, overweight from the 85th to under the 95th, obesity from the 95th, and severe obesity at "120% of the 95th percentile or greater, or 35 kg/m2 or greater".

Source 16

BMI misses some children's body fatStrong evidence

In 37 studies of 53,521 children and teenagers, the usual BMI cut-offs had a sensitivity of 0.73 and specificity of 0.93: they rarely flagged a child wrongly, but missed over a quarter of those with excess body fat.

Source 31

Measuring with care

NICE asks professionals to "think about the vulnerability of young people to eating disorders, and the impact of measuring their weight". If you're worried about a child's weight or eating, talk to their GP, school nurse or health visitor.

Sources 32, 14

Under 18

Adult ranges don't apply while you're growing. Use the NHS BMI calculator for children and teenagers (ages 2 to 17) in the UK, or the CDC's child and teen calculator in the US.

Sources 28, 14, 16

During pregnancy

The NHS calculator isn't for use in pregnancy, and the calculator stops if you tick that you're pregnant. NICE asks midwives to "measure the person's height and weight and calculate BMI at the first face-to-face antenatal appointment", and not to weigh you routinely after that without a clinical reason. Ask your midwife.

Sources 5, 33

If you have, or think you might have, an eating disorder

The NHS calculator isn't for you either. See a GP, or contact Beat: the support section below has the numbers.

Sources 5, 35, 36

A condition that affects your height

BMI depends on height, so a growth condition or a curved spine skews it. Clinicians can estimate height from the length of your forearm, your knee height or half your arm span instead; ask your GP or a dietitian.

Sources 5, 20, 34

After an amputation

Missing limbs lower your weight, so a standard BMI reads low. Clinicians adjust for it: one UK guide counts a whole leg as about 15.6% of body weight and a whole arm as 4.9%. Ask your clinical team or a dietitian.

Source 34

Fluid build-up

Swelling (oedema) or fluid in the abdomen (ascites), for example with heart, kidney or liver disease, adds weight that isn't fat or muscle. Clinicians judge weight differently then, sometimes using the size of the upper arm instead.

Source 34

Very muscular, or losing muscle

BMI can't tell muscle from fat. NICE's waist-to-height ratio bands can be used with high muscle mass, so use that calculator as well.

Source 1

"BMI measures how fat you are"Strong evidence

It measures weight for height. Across 25 studies, the usual cut-offs missed half the people with excess body fat, and the NHS says BMI "cannot tell the difference between muscle and fat".

Sources 8, 5

"Women, men and older people need different BMI ranges"Strong evidence

WHO's ranges are "age-independent and the same for both sexes", and the CDC uses them for adults regardless of age or sex. NICE asks for caution over 65 but gives no different ranges. What does differ is body fat at the same BMI: women carry more, and it rises with age. The only age-specific BMI cut-offs we found are for diagnosing malnutrition.

Sources 2, 27, 1, 17

"The healthy range is the same for everyone"Strong evidence

In the UK, NICE uses lower thresholds (23 and 27.5) for people of South Asian, Chinese, other Asian, Middle Eastern, Black African or African-Caribbean background. In England, BMIs with the same diabetes risk as 30 in White adults ranged from 23.9 (South Asian) to 28.1 (Black).

Sources 1, 7

"A healthy BMI means a healthy waist"Strong evidence

NICE: "a person in this category may nevertheless have central adiposity". In UK survey data, people with a healthy BMI but a waist-to-height ratio of 0.5 or more had worse risk factors than those with a smaller waist.

Sources 1, 25

"An overweight BMI means you're unhealthy", or "BMI means nothing"Strong evidence

Neither. On average, mortality rises gradually above 25 (7% higher at 25 to 27.5, 20% higher at 27.5 to 30) and more steeply above 30. For one person, BMI adds little to heart risk prediction once blood pressure, diabetes and cholesterol are known, and fitness shifts the risk a lot.

Sources 11, 12, 13

"BMI was designed to diagnose people"Strong evidence

Quetelet was a statistician describing "the average man". Keys chose the index for studies of whole populations because it was simple and could be applied "to all populations at all times".

Sources 18, 19

"Teenagers can use an adult BMI calculator"Strong evidence

The NHS says to use "the right calculator for adults (aged 18 and over) or children and teenagers (aged between 2 and 17)", and NICE asks for charts adjusted for age and sex.

Sources 28, 1

Help and support

If you have, or think you might have, an eating disorder, these numbers aren't the right tool: the NHS calculators say the same. Eating disorders can affect people at any weight. See a GP as soon as you can, or talk in confidence to the eating disorder charity Beat: 0808 801 0677 (England), 0808 801 0432 (Scotland), 0808 801 0433 (Wales) or 0808 801 0434 (Northern Ireland), or email help@beateatingdisorders.org.uk (Scotland, Wales and Northern Ireland have their own addresses on Beat's website, along with opening times). If you or someone else is in immediate danger, call 999 or Samaritans on 116 123; for medical advice, call 111.

Sources 5, 32, 35, 36

Signs to look out for

The NHS lists these among the signs of an eating disorder:

• Spending a lot of time worrying about your weight and body shape.
• Avoiding socialising when you think food will be involved.
• Eating very little food.
• Making yourself sick or taking laxatives after you eat.
• Exercising too much.
• Having very strict habits or routines around food.
• Changes in your mood, such as being withdrawn, anxious or depressed.

Source 35

Goals that aren't numbers

If numbers like these make you anxious, set goals for health and strength instead, such as walking further or lifting a little more (our suggestion, as in our exercise for weight loss guide).