Is there really an ideal weight?
No. The four clinical formulas disagree by around three kilograms at the same height, and none was designed as a health target — they were built to calculate drug doses. A healthy BMI range spans about twenty kilograms.
Updated 2026-08-24
The formulas were built for something else
Four equations turn up in every ideal weight calculator: Devine from 1974, Hamwi from 1964, and Robinson and Miller, both 1983. All four take height and sex and return a single number in kilograms.
None of them was derived from health outcomes. Devine, the most widely used, was published to standardise drug dosing — several medications are dosed on lean body mass rather than actual weight, and clinicians needed a consistent way to estimate it for a given height. It was an approximation invented for a practical clinical problem, and Devine himself offered no data to support it as a health target.
Hamwi came from a method for estimating calorie needs in diabetic patients. Robinson and Miller were later attempts to fit the same shape of equation to different reference populations. The lineage is clinical convenience throughout, and at no point does it pass through "what weight is best for a person to be".
That matters because the number gets quoted as though it were a goal. A figure designed to make a drug dose reproducible is being read as a verdict on someone's body, and the two have almost nothing to do with each other.
They do not agree with each other
Run all four at 175 centimetres for a male body and they return 70.5, 68.9, 68.7 and 72.0 kilograms — Devine, Robinson, Miller and Hamwi respectively. A spread of 3.3 kilograms.
That disagreement is not a rounding artefact; it is structural. Each formula uses a different base weight at five feet and a different increment per inch above it — Devine adds 2.3 kg per inch, Hamwi 2.7, Miller only 1.41. The taller you are, the further apart they get, because the increments diverge with every inch.
Presenting any one of them as "your ideal weight" therefore requires picking a winner among four equally arbitrary options, and calculators that show a single figure have quietly done exactly that. Showing all four is the honest presentation, because the disagreement is the finding.
They also share a structural blind spot: every one takes only height and sex. Frame size, muscle mass, limb proportions and age are all absent. Two people of the same height and sex get the same number regardless of build, which is precisely the criticism usually levelled at BMI.
A range beats a number
The more defensible answer is a range, and the usual source is BMI. A BMI between 18.5 and 24.9 at 1.75 metres corresponds to roughly 56.7 to 76.3 kilograms.
That is a span of nearly twenty kilograms, and the width is the point rather than a weakness. Healthy weight genuinely varies that much between people of the same height, because it depends on how much of that weight is muscle, how the skeleton is built, and what someone's history is.
BMI has its own well-documented problems — it cannot distinguish muscle from fat, it was derived from nineteenth-century European populations, and it misclassifies muscular people routinely. As a population screening tool it works; as an individual verdict it is weak. But even a weak range is more honest than a precise-looking single number derived from a drug-dosing convention.
Where a single figure is genuinely needed — a starting point for a calorie target, say — take the midpoint of the range rather than any one formula, and treat it as a hypothesis to be revised once you see how your body responds.
What actually predicts health
If the question behind "what should I weigh" is "am I healthy", weight is a poor way to ask it, and several better measures are just as easy to obtain.
Waist circumference is the strongest simple one. Fat carried around the abdomen is metabolically active in a way that fat on hips and thighs is not, and abdominal circumference predicts cardiovascular and metabolic risk better than total weight does. Waist-to-height ratio is the usual expression, with keeping your waist under half your height a common and defensible rule of thumb.
Body composition matters more than mass. Two people at identical weight and height with substantially different body fat percentages are not in the same position, and the scale reports them identically. This is the same argument that makes BMI weak, and it applies with equal force to any ideal weight formula.
And the things that most reliably track with health outcomes are behaviours rather than measurements — cardiorespiratory fitness, strength, sleep, and how much of your diet is unprocessed. Fitness in particular is a strong independent predictor: someone active at a higher weight frequently has better outcomes than someone sedentary at a "correct" one, which is difficult to reconcile with any single target number.
Using the number sensibly
None of this makes the formulas useless — it makes them a starting point rather than a destination, and the distinction changes how you should read the output.
Treat the four figures as a rough neighbourhood. If they cluster around seventy kilograms and you are ninety-five, that gap is informative regardless of which formula you prefer, and the disagreement between them is irrelevant at that distance. If you are seventy-three, the gap is smaller than the disagreement between the formulas and means nothing at all.
Be especially cautious in the other direction. These equations return low numbers for shorter people and can produce targets below a healthy BMI, which makes them actively unhelpful as goals for anyone already lean. Cross-check any target against the BMI range before treating it as something to aim at.
The most useful framing is that there is no ideal weight, only a range of weights compatible with health, and a set of behaviours that determine where in that range you comfortably sit. A calculator can show you the range. It cannot tell you what your body should be, and any tool that presents one number as though it could is overstating what the arithmetic supports.