How this instrument works
Height-for-age percentile tells you how a child's current height compares with a large reference population of children the same age and sex — nothing more. A boy at the 80th percentile for height is taller than roughly 80 out of 100 boys his age in that reference group. Height is one of the most heritable traits in pediatrics: tall or short parents tend to have children who track a correspondingly tall or short percentile for years without any underlying problem, which is why a single percentile reading rarely means much on its own.
This tool pulls its numbers from the CDC 2000 stature-for-age growth chart covering ages 2 through 20, built with the LMS statistical method (Cole 1990). At each reference age the chart stores three curve-shape numbers — L for skewness, M for the median height, and S for spread — and this calculator turns an entered height into a z-score against those three numbers before translating that z-score into a percentile with a logistic curve, the identical approximation formula this site's own IQ percentile tool already runs.
Keep two limits in mind. This page stores only 10 sample ages across the 2-to-20-year span rather than the CDC file's full month-by-month table, so it snaps your entered age to whichever of those 10 points sits closest rather than blending between them — an age that falls between two stored points is scored as though it were the nearer one, not averaged. And a height percentile is a screening signal, never a diagnosis: only a pediatrician reviewing the child's full growth curve, family heights, and pubertal stage can say whether an unusual or shifting number needs follow-up.
Height percentile also behaves differently from weight percentile in one important way: a healthy child's percentile band for height is expected to stay fairly steady year over year (short of the normal jump around a growth spurt), so a height percentile that drifts downward across two or more well-child visits is generally taken more seriously by pediatricians than a single low reading, since it can suggest slowing growth velocity rather than simply being a naturally short child.
- Choose Boy or Girl — the reference curves are sex-specific.
- Enter the child's age in months (24 to 240, i.e. 2 to 20 years).
- Enter the child's height in centimeters, measured standing without shoes.
- Read the percentile, the z-score, and the reference median height the calculator matched against.
- Bring an unusual result, or a flattening growth trend, to a pediatrician rather than acting on it alone.
Worked example — three heights checked against the CDC curve
Take a 10-year-old girl (120.5 months) measured at exactly 138.211 cm. That figure is precisely the CDC 2000 reference median height for a 10-year-old girl, so her z-score comes out to 0.00 and she sits precisely on the 50th percentile — the exact middle of the reference distribution by construction.
Now take an 8-year-old boy (96.5 months) measured at 122.0 cm, well under the 128.124 cm median height for boys that age. Running the LMS formula gives a z-score near −1.07, which the logistic step turns into roughly the 14th percentile — shorter than most 8-year-old boys in the reference group, a number worth flagging at his next visit alongside how his height has tracked over prior checkups.
Finally, a 14-year-old girl (168.5 months) at 168.0 cm compares against a 160.478 cm median for that age, producing a z-score around 1.14 and landing near the 87th percentile — taller than most girls her age in the reference population, plausibly explained by tall parents but still a data point a pediatrician would want in the context of her overall growth curve.
Questions
Does a short or tall height percentile need treatment by itself?
Not by itself. Plenty of healthy children sit consistently at the 5th or 95th percentile for height for their entire childhood, primarily because of parental height and genetics — a short percentile is not automatically a growth problem. The pattern pediatricians pay closer attention to is a height percentile that crosses two or more bands downward, or growth that visibly slows compared to a child's own established curve, since either can point toward an underlying issue worth investigating.
Why only 10 ages instead of a full month-by-month table?
The CDC's underlying stature-for-age file is published at fine age resolution, but fitting a full month-by-month lookup into this calculator's formula engine wasn't practical, so it stores 10 official reference ages spread across 2 to 20 years and snaps an entered age to whichever stored point is closest — no blending or estimating between points. If a child's exact age falls between two stored ages, expect the output to reflect the nearer one rather than a precisely interpolated value.
Will a pediatrician's chart match this calculator's number?
It should be close. Ages 2 through 20 are exactly the range where CDC recommends its 2000 growth charts as the U.S. reference for stature-for-age, and this calculator draws on that same official data file and LMS method a clinic's software or paper chart uses. Any small gap comes down to this tool's logistic-curve percentile shortcut and the age-snapping described above, not a different underlying dataset.
How precise is the percentile figure this tool reports?
The last step swaps in a logistic function for the true normal cumulative distribution function — a standard, well-known shortcut that this site's own IQ percentile calculator also relies on. Typical error against an exact table lookup runs around half a percentile point, which is plenty precise for a screening check but shouldn't replace a clinician plotting the same measurement directly on the official CDC chart.
Can a height percentile tell me if my child has a growth disorder?
No — a percentile from this calculator is a statistical comparison against a reference population, not a clinical diagnosis. Conditions like growth hormone deficiency, hypothyroidism, or a skeletal dysplasia are identified by a clinician evaluating growth velocity over multiple visits, family height patterns, pubertal timing, and sometimes bone-age X-rays or lab testing — never from a single height-percentile number, however unusual that number looks.
What is the reference median (M) figure shown in the results?
It's the CDC 2000 table's typical (50th-percentile) height for a child of the entered sex at the closest stored reference age. Seeing how far your child's height sits above or below that number, and in which direction, is a fast way to gauge the size and direction of the gap that produced the reported percentile.
References
- CDC — 2000 CDC Growth Charts Data Files (LMS parameters, STATAGE 2-20yr stature-for-age)
- CDC — Growth Charts overview: screening use and clinical interpretation
Read this first: This instrument computes a screening figure from population formulas — it is not a diagnosis, and it cannot see the whole picture a clinician can. Use it to inform a conversation, not to replace one.