Height Percentile Calculator

What a height percentile actually tells you

A height percentile answers one question: out of 100 people of the same sex (and, for children, the same age), how many are shorter than you? If you're at the 70th percentile, about 70 percent of the reference group is shorter and 30 percent is taller. That's all it means. A percentile is a description of position in a population, not a grade, and it says nothing about health on its own.

Percentiles are useful because raw heights are hard to interpret without context. "170 cm" is tall for a woman in most countries, short for a man in the Netherlands, and perfectly ordinary for a man in Japan. The calculator above handles three kinds of comparison: US adults (from measured national survey data), adults in about 200 other countries (from global estimates), and children and teens aged 2 to 20 (from CDC growth charts). This guide explains where each reference comes from, why the answers sometimes disagree, and how to read the result without over-reading it.

The US adult reference: NHANES 2021–2023

For US adults the calculator uses the National Health and Nutrition Examination Survey (NHANES), the federal survey that sends mobile examination centers around the country and measures a nationally representative sample. The numbers come from the 2025 report by Cheryl Fryar and colleagues at the National Center for Health Statistics, covering August 2021 to August 2023. It includes 2,690 men and 3,307 women aged 20 and over.

Two features make NHANES the best available US reference. First, heights are measured, not self-reported, by trained examiners using a stadiometer with shoes off, heels together and the head positioned in the Frankfort plane. Second, the sample is weighted to represent the whole civilian population, so the percentiles describe American adults as a whole rather than volunteers.

The report publishes the 5th, 10th, 15th, 25th, 50th, 75th, 85th, 90th and 95th percentiles. Between those points the calculator interpolates on the z-score scale, so its curve passes exactly through every published value. Beyond the 5th and 95th percentiles it extrapolates using the spread implied by the 5th–95th range: a standard deviation of about 7.6 cm for men and 7.0 cm for women.

Here is how some familiar heights rank among US adults aged 20 and over:

Notice how the tails compress. Going from 175 to 183 cm moves a man from the 49th to the 85th percentile, but the next 10 cm, to 193 cm, adds only 14 more percentile points. That's why z-scores, which keep counting standard deviations evenly, are more informative than percentiles for very tall or very short people.

Percentiles, z-scores and standard deviations

The calculator reports both a percentile and a z-score. A z-score counts how many standard deviations a height sits above or below the median. The two scales are linked through the normal distribution, and the table below translates between them for US adults aged 20 and over, using the same interpolation and extrapolation as the calculator.

Read across a row and you'll see why the sexes need separate references: a woman at +2 standard deviations is about the same height as a man just below his median. The rows at ±3 are extrapolations beyond the published data and should be treated as rough. In practical terms, the tallest 1 percent of US men start at about 192 cm (6′3.5″) and the tallest 1 percent of women at about 177 cm (5′9.7″).

Clinicians prefer z-scores for children at the extremes for the same reason. Two children at the "99.9th percentile" can be very different heights, while z-scores of +3.1 and +4.0 show the difference clearly.

A short history of height reference data

Height percentiles are only as current as the survey behind them. In the US, national reference charts for children were first published by the National Center for Health Statistics in 1977. They were revised in 2000 into the CDC growth charts that pediatricians still use, built from five national examination surveys plus supplementary data. For adults, NCHS has published anthropometric reference reports from successive NHANES cycles for decades. The 2025 edition used here covers measurements taken between August 2021 and August 2023, so it reflects today's adults rather than those of a generation ago.

Internationally, the WHO took a different route. Its child growth standards for ages 0 to 5, released in 2006, describe how healthy breastfed children should grow in good conditions, based on the Multicentre Growth Reference Study in Brazil, Ghana, India, Norway, Oman and the US. The CDC charts describe how US children did grow. That distinction between a standard and a reference is one reason US guidance switches from WHO to CDC charts at age 2.

The adult country estimates are the newest layer. Before NCD-RisC, global height comparisons often mixed measured and self-reported data from different years, which is why older "tallest countries" lists contradict each other.

Age changes the answer

NHANES also reports percentiles by decade of age, and the calculator lets you compare yourself with people your own age. The difference can be large.

Two effects are mixed together here. Older adults were born into earlier generations, which grew up with different nutrition and health care. And people lose some height with age as the spinal discs thin and posture changes. A cross-sectional survey can't separate the two, but the practical consequence is clear: a 183 cm man ranks around the 80th percentile among men in their 20s and around the 93rd among men in their 70s. If you want to know how you compare with your peers, choose your age group; if you want a single national figure, use "all adults."

Other countries: estimates from NCD-RisC

Very few countries publish measured height percentiles, so for the rest of the world the calculator uses the NCD Risk Factor Collaboration (NCD-RisC). Its 2016 analysis in eLife pooled 1,472 population-based studies with measured heights from more than 18.6 million people, deliberately excluding self-reported heights because of systematic bias, and estimated the mean height at age 18 for every birth cohort from 1896 to 1996 in 200 countries.

NCD-RisC publishes means, not percentiles. To turn a mean into a percentile, the calculator assumes heights follow a normal distribution with the same spread as US adults (SD about 7.6 cm for men, 7.0 cm for women). That's a reasonable approximation, but it's an estimate, and the result is labeled as such.

The same height ranks very differently from country to country:

For the full set of averages and their century-long trends, see average height by country, or jump to the Netherlands and Japan.

Why the US appears twice. The US NHANES option and the NCD-RisC estimate for the United States don't agree exactly. NHANES gives a mean of 175.1 cm for all men aged 20 and over measured in 2021–2023; NCD-RisC estimates 177.1 cm for men born in 1996, measured or modeled at age 18. They describe different groups (all adult ages versus one young birth cohort) using different methods (direct percentiles versus a modeled mean). For a US resident, the NHANES option is the better comparison.

Children and teens: the CDC growth charts

For ages 2 to 20, the calculator uses the CDC 2000 stature-for-age charts. According to the 2002 methods report by Robert Kuczmarski and colleagues, these charts were built mainly from national health examination surveys conducted between the 1960s and 1990s and smoothed into percentile curves.

CDC publishes the curves as three numbers for every half-month of age, using Tim Cole's LMS method:

• M is the median height at that age.
• S is the coefficient of variation, a measure of spread relative to the median.
• L is a power transformation that corrects for skewness.

A child's height X converts to a z-score with Z = ((X ÷ M)^L − 1) ÷ (L × S), and the z-score converts to a percentile through the normal distribution. The calculator interpolates L, M and S to the child's exact age in months, so a child of 9 years and 11 months isn't treated as a 9-year-old.

Two examples: a 10-year-old boy who is 140 cm tall has a z-score of about +0.21, around the 58th percentile, against a median of 138.6 cm. An 8-year-old girl who is 122 cm has a z-score of about −0.98, around the 16th percentile, against a median of 127.6 cm.

Why CDC rather than WHO? For children under 2, US health authorities recommend the WHO growth standards, which describe healthy breastfed infants from six countries. From age 2, the CDC's 2010 guidance in MMWR recommends the CDC charts, partly because measurement switches from lying length to standing height at that age and the CDC charts extend to age 20. The WHO also publishes a reference for ages 5 to 19, released in 2007 as a reconstruction of the 1977 NCHS/WHO reference. For school-age children in the US the two give similar pictures; the calculator uses CDC for consistency with US practice.

Reading your result sensibly

A few rules of thumb help put any percentile in perspective.

1. Percentiles describe, they don't judge. The 10th percentile is as normal as the 90th; by definition, 10 percent of healthy people are shorter.
2. Extreme values are approximate. Above the 95th or below the 5th percentile the calculator is extrapolating, so treat "99.7th" as "very tall" rather than a precise rank.
3. For children, trends matter more than snapshots. CDC guidance flags measurements beyond about two standard deviations (the 2.3rd and 97.7th percentiles) as worth a look, but a child who stays steadily on a low curve is usually just a small child. Crossing several percentile lines between check-ups is more meaningful. Plot repeat measurements on the growth chart to see the pattern.
4. Population averages shift. Our World in Data notes that men and women worldwide gained about 8 to 9 cm in average height over the last century, so the same height ranks differently depending on which generation and which country you compare with.

Measuring yourself for an accurate percentile

Because every 1 cm can move you several percentile points near the middle of the distribution, the measurement matters more than the math.

• Go barefoot and stand on a hard floor with your heels, buttocks, shoulder blades and the back of your head against a wall.
• Level your head so you're looking straight ahead, as in the NHANES protocol, and take a deep breath before marking.
• Use a rigid marker, such as a hardback book held flat against the wall, rather than a hand.
• Measure in the morning. A 1986 study in the British Journal of Sports Medicine measured an average daily height change of 14.4 mm, enough to shift a man near the median by around 7 percentile points.

Once you have your percentile, you can see what it looks like next to other people in the height comparison chart, or estimate a child's adult height with the height predictor.