How to read a growth chart
The page of curves in the health record is the densest piece of information in routine child health, and the most frequently misread. Understanding it saves a lot of unnecessary worry, and catches the real problems earlier.
Updated 19 September 20264 min read
A percentile is not a score
This is the most common misunderstanding. The 50th percentile is not a pass mark and not a target. It is the median: half of children of the same age and sex are above it and half below.
If your child sits at the 25th percentile, roughly 25 out of 100 same-age children weigh or measure less, and 75 measure more. That is entirely within the normal range. So is the 15th, and so is the 85th.
WHO defines 'outside the reference range' as a z-score beyond plus or minus 2, which corresponds to roughly the 2.3rd and 97.7th percentiles. By construction, about 5% of a healthy population falls outside those lines. Being outside them warrants a closer look. It does not by itself mean something is wrong.
The trend matters, not the position
If you remember one thing, remember this: a single measurement has almost no diagnostic value. A sequence of measurements does.
A child who has sat at the 15th percentile all along is most likely simply a small child, and perfectly healthy. A child who has slid from the 75th to the 25th is worth investigating even while still 'in the normal range', because they have left the track they were on.
That is precisely why every check-up takes a measurement and plots it on the same chart. An isolated number tells a doctor very little either.
Crossing two major percentile lines is generally a signal to assess. There is one exception under age two: birth size is heavily influenced by intrauterine factors, and many children shift towards their genetically determined track during the first two years, producing a sizeable but normal correction.
WHO standards versus national references
These two kinds of data are often used interchangeably, but they are not the same kind of thing.
The WHO Child Growth Standards (2006) are a standard. They were built from children in six countries raised under optimal feeding and health conditions, and they answer the question of how children should grow.
National growth surveys are references. They describe how children in that population actually grow, including whatever the real nutritional and health situation happens to be.
The two differ, particularly in infancy. The WHO sample was predominantly breastfed, so median weight between 3 and 12 months runs slightly lower than in mixed-feeding reference populations. Formula-fed babies can therefore look heavy against WHO standards. That is not a flaw in the standard; it is the standard doing its job of flagging a feeding difference.
This site uses WHO data: the 2006 Child Growth Standards for 0-5 years, and the 2007 Growth Reference for 5-6 years.
What each measure tells you
Weight-for-age: the most sensitive and the most volatile. A bout of diarrhoea or a cold can drop it. Useful for short-term nutritional status.
Height-for-age: reflects longer-term nutrition and health. Height changes slowly, so a shortfall usually means the problem has been running for a while.
BMI or weight-for-height: proportion rather than size. Weight at the 90th percentile means little on its own — if height is also at the 90th, that is simply a large child with normal proportions.
Head circumference: most important from 0-2 years, and a proxy for brain growth. Growth that is too fast or too slow needs immediate assessment. An abnormality here is far more urgent than an abnormality in weight.
In practice: read height and weight together, then look at BMI. That beats staring at any single line.
The 0.7cm step at 24 months
If you plot the numbers yourself, you may find your child 'shrinks' by about 0.7cm around their second birthday. Nothing was measured wrong.
Before 24 months the measurement is recumbent length. From 24 months it is standing height. The same child measures about 0.7cm taller lying down, because the spine compresses slightly under gravity.
WHO keeps this step in its tables and so does the chart on this site. So if you measured at 23 months and again at 25, check which position was used before concluding that growth has stalled.
When it is genuinely worth worrying
Do not wait for the next routine check if:
Weight or height z-score is below -2 (under the 3rd percentile) or above 2 (over the 97th).
Two or more successive measurements have crossed two major percentile lines.
Weight is static or falling for more than a month, under age two.
Head circumference growth clearly departs from its curve, in either direction.
Height velocity slows abruptly — a preschooler should gain roughly 5-7cm a year.
Any of the above comes with other signs: low energy, repeated infections, missed developmental milestones, a marked drop in appetite.
Conversely, a child growing steadily along their own curve, alert and meeting milestones, is very probably fine even at the 10th percentile.
This is general reference material. It does not replace advice from your doctor. Sources: WHO Child Growth Standards, China Dietary Guidelines (2022), China National Immunization Program schedule (2021), CDC developmental milestones.
Back to guides