A height percentile ranks a child against other children of the same age and sex in a reference population — it isn’t a grade or a target. The 75th percentile means taller than roughly 75 out of 100 children in that group, and shorter than the rest. A single reading matters far less than a steady trend over time, and any specific concern is best checked with a GP or paediatrician.
What a percentile actually means
A percentile is a ranking, nothing more. It compares one child’s height against a reference population of children the same age and sex, and reports where that height falls in the spread.
| Percentile | Plain-English meaning |
|---|---|
| 5th | Shorter than about 95 in 100 peers |
| 25th | Shorter than about 75 in 100 peers |
| 50th (median) | The typical, middle height for that age and sex |
| 75th | Taller than about 75 in 100 peers |
| 95th | Taller than about 95 in 100 peers |
None of these bands is “good” or “bad”. They simply describe where a height sits among many other, entirely normal, heights.
Percentile is also not the same idea as average. Average height at a given age is a single point — the median, or 50th percentile. Percentile describes where any specific height falls relative to the whole spread of heights around that point, not just the midpoint itself.
Reference populations like WHO 2007 and CDC/NHANES are built by measuring large numbers of real children at each age, then ranking those measurements to build the curve a percentile is read from. A percentile only ever describes a comparison against that particular measured group, at that particular age.
A worked example: an 8-year-old boy at 131.4cm
This site’s own CDC/NHANES reference table gives 131.4cm as the median height for an 8-year-old boy in the US sample. A boy at exactly that height sits at the 50th percentile: taller than roughly half of 8-year-old boys in that reference group, and shorter than the other half.
Try it yourself in the calculator below, pre-filled with this exact example.
Percentile
50th
At 131.4 cm and age 8, you are taller than approximately 50% of the CDC reference population for your sex.
z-score 0.00
A 50th-percentile reading isn’t a special achievement, and it isn’t average in the sense of “correct” either. It’s simply the midpoint of a wide, ordinary range of heights for that age.
Why the same height gets a different percentile on different charts
This site’s calculator offers two reference systems, and they can give a noticeably different answer for the same child. Take a 10-year-old boy at 140cm. Against the WHO 2007 international reference, that height works out at roughly the 63rd percentile. Against the CDC/NHANES US reference, the same 140cm works out at roughly the 33rd percentile.
That’s a 30-point gap for one unchanged height. It happens because WHO 2007 and CDC/NHANES are built from different reference populations, using different underlying samples and methods, not because one chart is “wrong”. WHO’s reference covers ages 5-19 as an international benchmark; CDC’s is drawn from a US sample and also covers adults.
Neither system is the single correct answer. If you’re comparing a child specifically against other children in the US, CDC is the closer match. For a broader international comparison, WHO is the more commonly used field standard. Picking a reference matters as much as the height itself.
| Name | Height |
|---|---|
| 8yo boy, US median (CDC) | 131.4 cm |
| 8yo boy, international median (WHO) | 127.7 cm |
A trend matters more than one reading
Clinicians generally care less about a single percentile reading than about the pattern over time. A child who tracks steadily along roughly the same percentile line for months or years is usually seen as reassuring, even if that line isn’t the 50th.
A sudden, significant shift is different. If a child’s height moves from tracking around, say, the 75th percentile down towards the 25th over a short period, that kind of percentile crossing is one of the reasons a paediatrician might look more closely. It doesn’t automatically mean a problem, but it’s a pattern worth a clinical opinion rather than a guess from a chart at home.
This is exactly why paediatric growth checks plot several visits over time, rather than relying on one measurement. A single number can’t show a trend; a series of them can.
What percentiles don’t mean
A percentile is not a health verdict. It doesn’t diagnose a condition, and a below-average reading, on its own, says nothing about a child’s wellbeing.
It’s also not a prediction of adult success, ability or anything beyond height itself. Height percentile is a purely descriptive statistic about one measurement against one reference group.
It isn’t something a child or parent should feel pressure about, either. Percentiles shift naturally with growth spurts and the timing of puberty, and a huge range of percentile positions is entirely normal and healthy.
These are population references, not individual medical assessments. A GP or paediatrician should evaluate any individual concern about a child’s growth, especially a sudden percentile crossing, rather than a chart alone.
Method and sources
The 8-year-old worked example uses this site’s CDC/NHANES stature-for-age table, where 131.4cm is the published median (50th percentile) for boys at that age. The WHO comparison figure (127.7cm) is the WHO 2007 reference’s median for boys at “8 years, 1 month”, the nearest whole-year point in this site’s table.
The 140cm, 10-year-old boy example uses the LMS method behind this site’s percentile calculator, applied to both the WHO 2007 (interpolated to age 10.0) and CDC/NHANES (age 10 exact) reference tables. Both tables, and the LMS maths itself, are documented in full in this site’s data and calculator source.
This is general reference information, not medical advice. If you have a specific concern about a child’s growth, a GP or paediatrician can assess it properly.
Read the average height for a 10-year-old to see the WHO/CDC gap in more detail, or try the full percentile calculator with your own child’s height and age. You can also open the comparison board to see any two heights stood side by side.