Is BMI Outdated? What the Criticism Gets Right
BMI is 194 years old, and Adolphe Quetelet built it in 1832 to describe populations. He said in the same work that it applied to groups and left individuals alone. In 1972 Ancel Keys tested it against every rival index of the day, found it the best available proxy for fatness in large samples, and gave it the name body mass index.
BMI does the job it was designed for. The rest of us keep asking it a question about one person, which it was never built to answer. Some of the complaints that follow from that are fair and some are lazy, and the two are worth separating.
What the criticism gets right
BMI measures how heavy you are for your height. That is all it sees. Your muscle, your bone density, your water and the location of your fat all arrive on the page as the same kilogram, so anybody else at your height and weight reads exactly as you do.
The gap is widest at the two ends. A trained lifter reads overweight while carrying single-digit body fat. An older adult who has shed muscle over twenty years can sit inside the healthy band while carrying more fat than they did at forty, because the muscle that left weighed something and the fat that replaced it weighs less per unit of volume.
Medicine has caught up with that. In June 2023 the American Medical Association adopted policy putting BMI alongside a direct measure of fat or fat distribution instead of leaving it to stand alone. In January 2025 the Lancet Commission on Clinical Obesity went further, asking that excess fat be confirmed by a second measurement before anyone receives an obesity diagnosis at all.
Height-based targets carry the same blind spot. The ideal weight calculator runs four formulas on your height and your sex, so it inherits the limitation. Treat its output as a reference figure and pair it with a direct measurement, which is what the AMA is asking for.
What the criticism gets wrong
Two objections come up constantly, and the evidence handles both.
The first says Quetelet was a mathematician, so the index is not medical. True, and irrelevant. Plenty of clinical tools were built by people outside medicine, and what matters is whether the tool predicts outcomes.
The second says BMI is useless. Across populations of tens of thousands it tracks body fat percentage closely enough to predict cardiovascular disease, type 2 diabetes and all-cause mortality. At the edges it fails on individuals. Both of those are true at the same time, and the second one is the one you live in.
Where the number still earns its place
BMI runs on two figures you already know, and it takes ten seconds. Every other measure of body composition asks you for money, equipment or both. For a population screen, or for a first look before you decide what else to measure, it is the right tool.
It is also a fair first pass on yourself, as long as you train casually, you are some way off old age, and your frame is ordinary. Run your height and weight through the BMI calculator and read the result as a starting point. The guide on what BMI measures covers where the formula comes from and the specific cases where it falls apart.
What to read alongside it
Two measurements close most of the gap, and both take under a minute.
Your waist divided by your height is the strongest of the cheap options. Under 0.5 is the target, and it responds to where your fat sits as well as how much of it there is. The piece on keeping your waist under half your height explains why that single ratio outperforms BMI for predicting metabolic risk, and the fat your BMI cannot see covers what that measurement picks up.
Then take a body fat estimate. A tape measure and the Navy circumference method get you within three or four points, which is enough to tell a lifter at 27 BMI apart from a desk worker at 27 BMI. The body fat calculator runs that method from three measurements.
Common questions
Has BMI been replaced by anything official?
No. The World Health Organization categories stand as they were, and BMI remains the entry point in most clinical guidelines. The instruction around it is the part that changed. Since 2023 from the AMA, and since 2025 from the Lancet Commission, the advice is to confirm with a second measure before you act on the number.
Is BMI accurate for women?
It is equally inaccurate for both sexes, in the same direction. Women carry more essential fat than men at any given BMI, so the same figure describes a higher body fat percentage in a woman. The categories were drawn once for everybody, which is one good reason to take a body fat estimate alongside yours.
Should I ignore my BMI if I lift weights?
Read it, then check it. If you train seriously and your BMI reads 26 to 29, a body fat measurement usually settles the question in a minute. Body fat in the fitness or athlete range means your BMI is describing your muscle, and you can leave it where it is.
Where that leaves the number
BMI is a population tool that has spent fifty years being used as a personal one, and the correction working its way through medicine is to pair it with something that can see fat. Take the number, take a waist measurement, and let the two of them disagree if they are going to.
Scope and correction note, 19 September 2026: the history, the AMA policy and the Lancet Commission recommendation are cited as published. Nothing on this page is medical advice. If a figure here is wrong, tell us and we will correct it and date the correction. See the editorial policy and medical disclaimer.