BMI for Women: What It Misses, and What to Use Instead

BMI for women uses exactly the same thresholds as for men: below 18.5 is underweight, 18.5–24.9 is the "healthy" range, 25–29.9 is overweight, and 30 or above is obesity. Those numbers are real and your clinician will use them — but they were never designed to describe one person. BMI is a population screening tool, invented in the 1830s from measurements of European men, and it cannot tell muscle from fat, cannot see where your fat is stored, and knows nothing about what menopause does to body composition. Waist-to-height ratio and waist circumference tell you more about your actual risk than BMI ever will.

First, the chart you came for

BMI is weight in kilograms divided by height in metres squared. If you want the arithmetic done for you, use our BMI calculator. These are the categories the CDC, the NHS and virtually every clinic in the world apply to adults 20 and over, regardless of sex or age.

And the other number people search for. In the CDC's measured national survey data (NHANES 2015–2018), the average US woman aged 20 and over is about 5 ft 3.5 in tall and weighs 170.8 lb (about 77.5 kg) — which works out to a BMI of roughly 29.8, the very top of the "overweight" band. Her average waist is 38.7 inches, already past the 35-inch risk threshold. An average is a description of a population, not a target for a person, and it is worth sitting with the fact that the "average" woman and the "healthy BMI" woman are now two different people. Neither figure tells you anything about the woman reading this; there is more on that in our piece on average weight for women.

Where BMI came from — and what it was for

The formula is not medical in origin. Adolphe Quetelet, a Belgian astronomer and statistician, published it in 1832 while trying to describe l'homme moyen — the average man — using measurements from European men. He was building a statistical portrait of a population. He was explicitly not measuring anyone's health, and he had no interest in body fat.

It sat mostly unused until 1972, when physiologist Ancel Keys tested several weight-for-height formulas against body fat in large male cohorts, found Quetelet's index the least bad of the simple options, and renamed it the "body mass index." Keys recommended it for population studies and warned against using it to judge individuals. Insurers and public-health agencies adopted it anyway, because it is free, instant, and needs nothing but a scale and a tape measure. That is the entire reason BMI is in your medical chart: not accuracy, but cheapness.

In 2023 the American Medical Association formally adopted policy stating that BMI alone is an imperfect clinical measure, that it does not distinguish fat from lean mass or capture where fat sits, and that it should be used alongside other measures such as waist circumference and body composition.

Five things BMI gets wrong for women

1. It cannot tell muscle from fat

BMI is a ratio of total mass to height. Muscle is roughly 18% denser than fat, so two women of the same height and weight — one who lifts, one who does not — get the identical BMI while having completely different bodies. A 5 ft 5 in woman who strength-trains seriously can land at a BMI of 26 with a body-fat percentage in the low 20s and flawless bloodwork, and be told she is "overweight." This is not a rounding error; it is the formula working exactly as designed and being asked the wrong question. If you have been reclassified upward after adding strength training, the number moved because you built tissue you want.

2. It says nothing about where fat sits

Fat is not one substance. Subcutaneous fat — hips, thighs, the layer you can pinch — is metabolically fairly quiet. Visceral fat, packed around the liver, pancreas and intestines, is not: it is lipolytically active, drains directly into the portal vein, and drives insulin resistance, atherogenic lipid changes and inflammation. Two women with a BMI of 27 can have wildly different visceral fat volumes and therefore wildly different cardiometabolic risk. BMI cannot see the difference. A tape measure around the waist can. This is the whole reason midlife waist changes matter more than the scale reading that accompanies them.

3. It performs differently across ethnicities

The thresholds were calibrated on European-ancestry populations. A 2004 WHO expert consultation, published in The Lancet, reviewed the evidence and concluded that people of Asian ancestry — particularly South Asian and East Asian — carry more body fat and more visceral fat at any given BMI, and that risk of type 2 diabetes and cardiovascular disease rises at BMIs well below 25. Rather than redefine the international cut-offs, WHO proposed a series of additional public-health action points along the BMI range — at 23.0, 27.5, 32.5 and 37.5 kg/m² — and many clinicians now treat 23 as the point at which a South Asian woman warrants metabolic screening. Applying a single 25 threshold to every woman on earth systematically under-detects risk in some groups and over-flags it in others.

4. It ignores the menopause transition

The SWAN study followed women through the menopause transition with repeated DEXA scans, and found something the scale hides: starting roughly two years before the final period, the rate of fat gain roughly doubles while lean mass begins to fall. A woman can hold her weight — and her BMI — perfectly steady across those years while quietly trading muscle for fat, and specifically for visceral fat as oestrogen falls. Her BMI says "nothing has changed." Her metabolism disagrees. This is why menopause weight gain so often feels like a redistribution rather than a gain, and why muscle loss in midlife is the more consequential story.

5. It is blind to "normal weight obesity"

This is the failure that matters most for midlife women, and almost no BMI page mentions it. Normal weight obesity means a BMI in the healthy 18.5–24.9 range alongside a high body-fat percentage — commonly taken as above about 30% in women, though the landmark study used a cut-off of 33.3%. In that NHANES-based analysis, published in the European Heart Journal in 2010, people with a normal BMI but high body fat had a markedly higher prevalence of metabolic syndrome, and among the women, significantly higher cardiovascular mortality than normal-BMI women with normal body fat. They were, by every chart on every clinic wall, "fine."

If you are a woman in your fifties whose weight has not changed since your thirties but whose clothes fit differently, whose waist has thickened, and who has lost strength — you may be exactly this person, and BMI will never find you.

What to use instead

Waist-to-height ratio (the single best free measure)

Measure your waist at the midpoint between the bottom rib and the top of the hip bone, at the end of a normal breath out, tape snug but not compressing. Divide by your height in the same units. NICE now recommends this alongside BMI for anyone with a BMI under 35 — of either sex, any ethnicity, including people with high muscle mass — because it captures central adiposity, the part that matters, and it self-corrects for body size.

The rule of thumb is simply: keep your waist under half your height. A 5 ft 4 in (64 in) woman is aiming for a waist under 32 in.

Waist circumference

The blunter version, and the one most clinicians know by heart: for women, a waist above 35 in (88 cm) marks increased risk of heart disease and type 2 diabetes, per NHLBI guidance. International diabetes criteria use a lower threshold of 31.5 in (80 cm) for central obesity in women. It ignores height, which is why waist-to-height ratio is better — but it takes ten seconds and it beats BMI.

DEXA and bioimpedance

A DEXA scan (the same machine used for bone density testing) is the practical gold standard for body composition: it reports fat mass, lean mass, and — on newer software — an estimate of visceral adipose tissue, region by region. If you have ever wondered whether you are the "normal BMI, high body fat" woman, this is the test that answers it. It costs money and typically is not covered for this purpose.

Bioimpedance (smart scales, handheld devices, in-gym analysers) passes a tiny current through the body and estimates fat from tissue resistance. Its absolute numbers can be off by several percentage points and drift with hydration, time of day, and where you are in your cycle. Its trend, measured under identical conditions each time, is still genuinely useful: if lean mass is falling over six months while weight holds, that is real information a scale cannot give you.

So is BMI useless?

No — and pretending otherwise is its own kind of dishonesty. Across large populations, BMI tracks the risk of type 2 diabetes, cardiovascular disease, several cancers and all-cause mortality remarkably well for a formula that needs no equipment. Epidemiology, drug dosing, surgical eligibility and public-health surveillance all lean on it for good reason. At the extremes — a BMI of 17, a BMI of 42 — it is rarely wrong about the direction of risk.

The failure is one of resolution, not of concept. BMI is a good instrument for describing ten thousand women and a poor instrument for judging one. Treat your number as a single, cheap data point that says "look closer" or "carry on" — never as a verdict on your health, your discipline or your body. Use the BMI calculator to get it, then measure your waist, and take the second number more seriously than the first.

When to see a doctor

Book an appointment if any of the following apply — regardless of what your BMI says:

• Your waist-to-height ratio is 0.5 or above, even with a "normal" BMI. Ask for fasting glucose or HbA1c, a lipid panel, blood pressure and, if relevant, liver enzymes.
• Unintentional weight loss — more than about 5% of body weight in 6–12 months without trying. This always warrants investigation.
• Unexplained weight gain, especially with fatigue, cold intolerance, hair thinning or constipation — this can point to thyroid disease, and other causes of unexplained gain deserve a look too.
• A BMI below 18.5, or weight loss driven by restriction, purging or exercise you feel unable to stop. Disordered eating is a medical condition, not a willpower problem, and it is treatable.
• You are of South Asian, East Asian, or South-East Asian ancestry with a BMI of 23 or above — ask for metabolic screening earlier than the standard thresholds would suggest.
• You have been told to lose weight without any assessment of your blood pressure, glucose, lipids, or waist. That is a chart reading, not a clinical evaluation. It is reasonable to ask, politely and directly, what measurements the advice is based on.

Never set a weight target from a chart alone, and never start or stop a medication because of a BMI number. What to do about a raised waist or an abnormal lab is a conversation with a clinician who knows your history — and a good one will look at far more than your BMI.

More on this: weight & metabolism, how metabolism really changes with age, and our weight range tool, which we would rather you treat as context than as a goal.