Why your growth chart doesn't match your doctor's

There are two sets of curves, and they answer different questions

The WHO Child Growth Standards (2006) come from the WHO Multicentre Growth Reference Study. It followed healthy children in Brazil, Ghana, India, Norway, Oman and the United States, raised in conditions that support healthy growth, including breastfeeding. Every infant in the reference group was breastfed for 12 months and predominantly breastfed for at least 4 months. WHO calls the result a standard: a description of how young children grow when those conditions are met. ^1,6

The CDC 2000 growth charts were built from US national survey data and other records collected over roughly three decades. They describe how a large sample of American children actually grew at that time, when formula feeding was more common: about half of the infants in the data set had ever been breastfed. CDC calls them a reference. ¹

Neither chart is "wrong". They are different measuring sticks, and the same child lands in a slightly different place on each.

CDC's rule for the US: WHO before 2, CDC after

Since 2010, CDC has recommended that clinicians in the United States use the WHO standards for children from birth to 2 years, whatever the baby is fed, and the CDC 2000 charts from 2 years on. The recommendation was developed with the National Institutes of Health and the American Academy of Pediatrics. ^1,2,3

The switch at 2 is practical as well as scientific. At about that age children are measured standing up instead of lying down, so a new chart is needed anyway, and after 2 the methods behind the two sets of curves are similar. ¹

Many parents have heard it the other way round. If a website or app gives you a different percentile from your pediatrician's, the first thing to check is which chart it used for that age. childow picks the standard from your child's age and prints which one it used above every chart.

How big is the gap?

It depends on age. These are the median weights for girls from both sets of published tables. The median is the 50th percentile: half of children weigh less, half weigh more.

Computed with the LMS method from the WHO tables published by CDC and from the CDC 2000 infant tables, both retrieved 2026-10-05. Boys show the same pattern.

In the first few months the WHO medians are heavier, so the same weight lands on a lower WHO percentile. Around 6 months the two nearly meet. After that the CDC 2000 medians are heavier, so the same weight lands on a higher WHO percentile. A 12-month-old girl exactly on the WHO median would sit at about the 28th percentile on CDC 2000.

This matches what CDC describes: healthy breastfed babies tend to gain faster than formula-fed babies in the first months and more slowly for the rest of the first year. A breastfed baby who looks as if she is "falling" across CDC 2000 curves after about 3 months may be tracking steadily on WHO. ¹

Six other reasons the numbers can differ

1. Different outer lines. On WHO charts, CDC recommends using the 2.3rd and 97.7th percentiles (two standard deviations, labelled 2nd and 98th on the printed charts) as the cutoffs, rather than the 5th and 95th that were traditional on CDC charts. A chart that looks "wider" may simply be drawn with different lines. ¹
2. Length versus height. Under 2, children are measured lying down (length). From 2, standing (height). Standing height reads a little shorter: WHO adds 0.7 cm to a standing height measured before age 2, and CDC subtracts 0.8 cm from a lying length measured at 2 or older. ⁴
3. Weight-for-age versus weight-for-length. Your doctor may be looking at weight relative to length, not to age. Those are separate charts with separate percentiles, and a long, lean baby can be low on one and average on the other.
4. Corrected age. For babies born early, AAP advises using corrected age for the first 2 years. A calculator that uses the birth date alone will put a preterm baby on a lower percentile. ⁵
5. The scale and the clothes. A clinic scale, an undressed baby and a clean diaper are not the same as a home bathroom scale after a feed.
6. Rounding and units. Entering "15 lb" for a baby who weighs 15 lb 12 oz is a difference of about 0.34 kg. For a 6-month-old girl on WHO that moves her from about the 28th to about the 43rd percentile.

How to compare two numbers fairly

If you want to understand a gap between your number and the one in the office, line up the details:

• Ask which chart the office used, and whether the number was weight-for-age or weight-for-length.
• Use the same date and the same measurement. A weight from a home scale two days later is a different data point.
• For a baby born before 37 weeks, check whether corrected age was used.
• Enter weight with its ounces or decimals. Whole pounds can hide a big share of a small baby's weight.

The growth chart on childow does this lining-up for you. Turn on Compare with my doctor's chart and it shows the same point on WHO and on CDC 2000 side by side, with a one-line explanation of the gap generated from fixed rules, not free-text AI.

What actually matters: the pattern

A percentile is a position, not a score. A healthy child can sit on the 10th percentile or the 90th. What pediatricians look at is the pattern across several visits, measured the same way on the same chart, together with how the child is eating, sleeping and developing. CDC's guidance stresses using several data points over time, alongside the child's medical and family history. ¹

It is reasonable to ask your pediatrician about a measurement that crosses two or more of the major percentile lines between visits, a point below the 2nd or above the 98th percentile on the WHO chart, or anything else that worries you. childow shows positions on reference curves. It does not diagnose, and it is not a substitute for your child's clinician.

Where the numbers on childow come from

Every percentile on this site is computed from the published LMS tables: the WHO tables for birth to 24 months as published by CDC, and the CDC 2000 tables from 2 to 20 years, with CDC's 2022 extended method for BMI above the 95th percentile. Each file is mirrored with a checksum, and our engine is tested against the percentile columns in the same files. The full list, with retrieval dates, is on the sources page.