Height Calculator

Three calculators sit above this. The first predicts a child's adult height from their own height and weight together with both parents' heights, using the Khamis-Roche method. The second uses the parents' heights alone, which works before a child is born or while they are still too young to measure usefully. The third converts a height between feet and inches and centimetres in either direction.

What decides how tall someone ends up

Height is mostly inherited. Estimates put the genetic contribution at somewhere between 60% and 80%, with the rest coming from environment, chiefly nutrition and health during childhood and during the mother's pregnancy.

Inherited does not mean copied. Height regresses toward the mean, which means very tall parents tend to have tall children who are usually a little shorter than they are, and very short parents tend to have short children who are usually a little taller. Each generation gets pulled back toward average, and that is why a prediction built from parental height alone is a range rather than a number.

It is also not one gene. Hundreds of common variants each nudge height slightly, which is what makes it behave like a smooth continuous trait rather than something that comes in a few discrete outcomes.

How growth actually runs

Stage	What growth does
Birth to about 2	Fastest of the whole of life, then slowing sharply
About 2 to puberty	Steady and fairly slow, a few centimetres a year
Puberty	A second peak, the growth spurt
After the spurt	Slows to a stop as the growth plates close
Late adulthood	Gradual loss of height, near universal in the very old

Two periods do most of the work. Infants grow faster than at any other point in life, and the rate falls away steeply to about age two. Growth then ticks along slowly until puberty brings a second peak, after which it tails off to nothing as the growth plates in the long bones fuse and close.

That closure is what makes the end point fixed. Growth trails off to zero at around 15 in girls and around 18 in boys on average, with wide individual variation, and once the plates have closed no amount of nutrition, exercise or sleep adds height.

At the other end of life height goes the other way. Intervertebral discs lose water and compress, posture changes, and vertebral bone can thin, so most people are measurably shorter in old age than at 25. It is close to universal in the very elderly.

Ways of predicting adult height

Method	What it needs	Roughly how close
Bone age (Greulich-Pyle)	A hand and wrist radiograph	The most accurate of the common methods
Khamis-Roche	Child's height, weight, age and both parents' heights	About 2.2 in for boys, 1.7 in for girls
Growth chart tracking	Repeated measurements over years	Good while the child holds a centile
Mid-parental height	Both parents' heights only	A range of about 8.5 cm either side
Doubling the height at age 2	One measurement	A rough guide, nothing more

None of these is exact. Height prediction is not an exact science whatever method you use, and a child can finish well outside what any of them said.

Mid-parental height

Average the two parents' heights, then add 2.5 inches for a boy or subtract 2.5 inches for a girl. That offset is roughly half the average adult difference between men and women, which is where it comes from rather than from anything about the individual child.

Clinicians usually quote a target range of about 8.5 cm either side of that figure, which covers most real outcomes. The range is the honest part of the answer. A prediction of 174 cm really means somewhere around 165 to 183 cm, and reporting the midpoint without the spread makes the method look far more precise than it is.

Bone age

Skeletal maturity predicts remaining growth better than age does, because two children of the same age can have very different amounts of growing left. The Greulich-Pyle method compares a radiograph of the left hand and wrist against a standard atlas, and the percentage of adult height already reached at that bone age gives the prediction.

It is the most accurate of the routine methods and it needs an X-ray, so it belongs in a clinic rather than on a website. One limitation is worth knowing: the atlas data were collected between 1931 and 1942 from Caucasian children in one American city, which limits how well it transfers to children today.

The Khamis-Roche method

This one uses the child's own stature and weight alongside the average of the parents, and it is considered among the more accurate methods that need no radiograph. For each half-year of age from 4.0 to 17.5, and separately for boys and girls, it applies four published coefficients:

adult height = B₀ + b₁·(child's height) + b₂·(child's weight) + b₃·(mid-parent height)

Its published standard error is about 2.2 inches for boys and 1.7 inches for girls, and it was developed on Caucasian children, with the method described as most applicable between ages 4 and 9 and to children free of growth-related conditions.

The first calculator above runs this method. It uses the 1994 published coefficients with the corrections from the 1995 erratum applied, since the weight coefficients in the original printing were wrong, and it shows which row of the table it used along with the equation, so the arithmetic is checkable rather than hidden.

Two details are worth knowing about how it behaves. The coefficients exist only at half-year steps, so an age of 5.2 is rounded to 5.0 before the lookup rather than interpolated, which is what the reference implementation does. And the result carries two published bands: half of children finish within about 0.9 inches of the prediction for boys and 0.7 for girls, and nine in ten finish within about 2.1 and 1.7 inches. Both are shown, because a prediction quoted to a tenth of an inch with no band around it claims a precision the method has never had.

If you compare the output against another site and see a difference of half an inch or so, the likely reason is the erratum. Calculators built from the 1994 printing without the correction return slightly different figures, and the corrected coefficients are the ones to trust.

Growth charts

The CDC growth charts plot height, weight and head circumference against age as a set of centile curves. Their value is in the tracking rather than in any single reading. Children tend to hold roughly the same centile through childhood, so a child sitting on the 40th line at four and still on it at eight is likely to finish near the 40th centile of adult height.

What matters clinically is a child crossing centiles, upward or downward, because that is a change in the pattern rather than a position within it.

The simple rules

Two shortcuts circulate widely. One is the parental average plus or minus 2.5 inches, which is the mid-parental method above. The other is doubling the height reached at age two for a boy, or at 18 months for a girl. They are quick, they are memorable, and they are less accurate than anything else here. Treat them as conversation rather than as planning.

Can anything actually make a child taller?

Not beyond their genetic potential, and nothing at all once the growth plates have closed. What good conditions do is prevent a child from falling short of that potential, which is a real effect and worth having.

1. Eat mostly unprocessed food. Fresh fruit and vegetables, whole grains, protein and dairy. Chronic undernutrition during childhood is the environmental factor with the clearest effect on final height, and it is largely why average heights in many countries rose sharply over the twentieth century.
2. Go easy on sugar, trans fats, saturated fat and salt. Not because they stunt growth directly, but because what they displace from the diet does.
3. Exercise regularly. It builds bone and muscle and supports general health. It does not stretch anyone, and no sport makes a child taller, whatever the correlation between basketball and height suggests about who takes it up.
4. Mind posture. Slouching costs visible height immediately, and over years a habitual curve can become a structural one.
5. Sleep properly. Growth hormone is released mainly during deep sleep, so a chronically short night through adolescence is one of the few controllable things that can plausibly cost a child height.

What is not on that list is any supplement, stretch or device sold on the promise of extra inches. Where a medical condition is limiting growth, a doctor can sometimes treat it, which is a different matter entirely and one worth raising early rather than late.

When a prediction is worth a doctor rather than a calculator

A few patterns are worth asking about: a child crossing downward through centile lines over a year or two, a child much shorter or taller than the mid-parental range predicts, no growth spurt by an age when peers have had one, or a growth rate that has clearly slowed while the child is still young. None of these means something is wrong, and all of them are better checked than watched.

Reading your result

Take the range seriously and the midpoint lightly. The single figure is the centre of a distribution roughly 17 cm wide, and half of all children land outside the middle of it.

Enter each parent's height as accurately as you can, since an error of an inch in either parent moves the prediction by half an inch. Measured heights beat remembered ones, and adults often report themselves an inch taller than they measure.

For weight against height rather than height alone, use the BMI Calculator, and the Ideal Weight Calculator works from an adult height to the usual healthy weight ranges.