How Tall Will I Be? Predicting Adult Height Explained
How tall will I be as an adult? A full walkthrough of the mid-parental formula and a WHO growth-based full predictor, with a worked example.
"How tall will I be?" is one of those questions almost everyone asks at some point, usually as a child sizing up their parents, or as a parent watching a growth spurt and wondering where it will end. There is no way to answer it with certainty, but there are several honest, well-documented ways to make a reasonable estimate, and this guide walks through the two this site's height calculator uses, plus the wider set of methods researchers and pediatricians actually rely on.
How tall will I be? Why the question is genuinely hard to answer
Adult height is the end result of a long, uneven process, not a fixed number waiting to be read off a chart. It depends on genetics inherited from both parents, but also on nutrition, general health, sleep and the exact timing of puberty, all interacting over roughly two decades of growth. Two children with identical parents can end up different heights; the same child measured at two different ages can look like they are tracking toward two different adult heights, depending on exactly where they are in their own growth timeline. Any calculator, this site's included, is working with partial information and returning a reasonable estimate, not a fact about the future.
The genetics behind height
Twin and family studies consistently estimate the heritability of adult height at around 60-80%. That figure describes how much of the height variation between people in a population statistically lines up with genetic variation between them, not how much of any one person's height is "genetic" versus "environmental": an individual's height is not neatly divisible that way. Height is also polygenic: genome-wide studies have linked hundreds of individual genes to height, each contributing a small effect, rather than one or two genes doing most of the work. That is a large part of why siblings raised in the same household by the same two parents can still end up noticeably different heights.
There is a second, subtler genetic-statistical effect worth knowing: regression to the mean. On average, children of unusually tall parents tend to land somewhat closer to the population average than their parents did, and children of unusually short parents likewise tend to land closer to average than their parents. This is not evidence against heritability, it is exactly what happens whenever a trait depends on many small, largely independent factors: the rare combination of factors that produced an extreme parent is less likely to be inherited in full by the next generation. The concept has a well-known origin story: the Victorian polymath Francis Galton named it "regression toward mediocrity" in the 1880s, while studying, appropriately enough, how parents' heights related to their children's. The modern statistical term "regression," used constantly across every field of data analysis today, traces directly back to Galton's height research.
Modern genetics has taken this much further than Galton could have. Large genome-wide association studies have now linked several thousand individual genetic variants to adult height, and research-grade "polygenic scores" built by combining all of them can explain a substantial share of height variation within a population, in some recent studies approaching the ceiling suggested by twin-study heritability estimates. That kind of DNA-based prediction is still a research tool rather than something available through a consumer calculator, and it estimates a range around a population, not a guaranteed number for one individual, but it is worth knowing that "genetics predicts height" is no longer just a twin-study statistic; it is now, increasingly, a specific, countable list of genetic variants.
Method 1: the mid-parental height formula
The older and simpler of the two methods, mid-parental height, has circulated in pediatric reference material for decades. It averages both parents' heights and applies a fixed adjustment for the child's sex, reflecting the average height gap between adult men and women:
- Boy: (mother's height + father's height + 13) ÷ 2
- Girl: (mother's height + father's height − 13) ÷ 2
Both formulas are in centimetres, and the result carries a commonly cited expected range of about ±8.5cm, reflecting how much individual children vary even when their mid-parental number is identical.
Worked example. Mother 160cm, father 175cm, girl: (160 + 175 − 13) ÷ 2 = 322 ÷ 2 = 161.0cm, with an expected range of about 152.5cm to 169.5cm.
Method 2: the current-growth trajectory, and the WHO table behind it
The height calculator's Full Predictor tool adds a second, independent estimate: instead of asking what the parents suggest, it asks how much of adult height a typical child of this exact age and sex has already reached, then scales this specific child's current height up by that same proportion.
| Age | Boys | Girls |
|---|---|---|
| Birth | 28.3% | 30.1% |
| 6 months | 38.3% | 40.3% |
| 1 year | 42.9% | 45.4% |
| 2 years | 49.7% | 53.0% |
| 3 years | 54.4% | 58.3% |
| 4 years | 58.5% | 63.0% |
| 5 years | 62.3% | 67.1% |
| 6 years | 65.7% | 70.6% |
| 7 years | 68.9% | 74.0% |
| 8 years | 72.1% | 77.6% |
| 9 years | 75.1% | 81.2% |
| Age | Boys | Girls |
|---|---|---|
| 10 years | 78.0% | 85.0% |
| 11 years | 81.1% | 88.9% |
| 12 years | 84.4% | 92.7% |
| 13 years | 88.4% | 95.8% |
| 14 years | 92.4% | 98.0% |
| 15 years | 95.7% | 99.1% |
| 16 years | 97.9% | 99.6% |
| 17 years | 99.2% | 99.8% |
| 18 years | 99.8% | 99.9% |
| 19 years | 100.0% | 100.0% |
Where this table comes from. The underlying heights are official WHO median length/height-for-age figures: the WHO Child Growth Standards for birth to 5 years, and the WHO Growth Reference 2007 for 5 to 19 years. The percentage column itself is not something WHO publishes directly, it is this table's own calculation, each age's median height divided by the median height at 19 years (treated as adult height), cross-checked against the growth progression reported in the Merck Manual's consumer growth-chart reference and a 2024 peer-reviewed longitudinal cohort study, for consistency rather than as a second independent source.
A naming mix-up worth avoiding. Some competing calculators label a table like this one "Bayley-Pinneau." That name belongs to a different, older method (Bayley N, Pinneau SR, "Length of maturation and prediction of adult height," Journal of Pediatrics, 1952) that predicts adult height from a child's bone age, read from a hand and wrist X-ray, not from calendar age. A bone-age table and a calendar-age table are not interchangeable. Every figure above uses calendar age only; WHO growth data, not Bayley-Pinneau, is this table's real basis.
A fractional age is interpolated, not looked up as a single row. An age between two rows (7 years 4 months, say) is estimated by drawing a straight line between the two nearest rows and reading off the point 4/12 of the way along it.
A fully worked example, start to finish
Take a girl aged 7 years 4 months, currently 122cm tall, with a mother who is 160cm and a father who is 175cm.
Step 1: interpolate the percentage. 7 years 4 months is 7.333 years. Girls are 74.0% of adult height at 7 years and 77.6% at 8 years, a gap of 3.6 percentage points. 4 months is 4/12 = 0.333 of the way from 7 to 8, so: 74.0 + (0.333 × 3.6) = 74.0 + 1.2 = 75.2%.
Step 2: scale her current height up by that percentage. 122 ÷ (75.2 ÷ 100) = 122 ÷ 0.752 = 162.2cm. This is her current-growth estimate.
Step 3: the mid-parental estimate, for comparison. (160 + 175 − 13) ÷ 2 = 161.0cm, exactly as worked through in the previous section.
Step 4: average the two. (161.0 + 162.2) ÷ 2 = 323.2 ÷ 2 = 161.6cm. That is the Full Predictor's headline number for this child, built transparently from two independently computed estimates rather than one formula alone.
Reading a growth chart: percentiles and what they imply
A growth chart plots height against age, with curved lines marking percentiles. A child on the 50th percentile is taller than half of the children of the same age and sex. For prediction, the number matters less than whether the child follows one curve over time. Children who stay near the same percentile usually finish near the matching adult percentile, which is one reason a pediatrician measures at every visit. A child who drifts across curves, upward or downward, is not necessarily unhealthy, because growth timing varies: some children spurt early and finish early, while others spurt late and keep growing after classmates have stopped. Puberty timing explains a large part of the difference, so a 13-year-old's current height is a weaker clue than their height plus the stage of puberty they are in. The growth-based predictor uses age and height together for exactly this reason.
How this compares with other prediction methods
The two methods above are deliberately simple and require nothing beyond a birthdate, a tape measure, and the parents' heights. Clinical and research settings have more demanding, and generally more accurate, alternatives:
- Bone-age methods (Greulich-Pyle). A hand and wrist X-ray is compared against a published reference atlas (Greulich WW, Pyle SI, Radiographic Atlas of Skeletal Development of the Hand and Wrist, 1959) to read a child's bone age, which can run ahead of or behind their calendar age. A child with "advanced" bone age typically has less growing left than their birthdate alone would suggest, and vice versa.
- Khamis-Roche. A statistically calibrated regression method (Khamis HJ, Roche AF, "Predicting adult stature without using skeletal age: the Khamis-Roche method," Pediatrics, 1994) built from the long-running Fels Longitudinal Study. It combines a child's current height, weight, age and mid-parental height without needing an X-ray, and is generally regarded as one of the more accurate non-X-ray methods for children roughly 4 to 17 years old. It is a genuinely different, more heavily calibrated calculation than the simple averaging this site's Full Predictor uses.
- CDC growth charts. Rather than predicting a single adult number, these track a child's height percentile against a large reference population across repeat visits. Consistency of percentile over time, not any single measurement, is the pattern pediatricians actually watch for.
None of these replace a pediatrician's judgment, and this site does not attempt to reproduce the X-ray-based or fully calibrated regression methods. The two tools on the height calculator are transparent, general-interest estimates, clearly labelled as such.
Rules of thumb people use, and what they ignore
Several shortcuts circulate online. The most common is doubling a boy's height at age 2, or a girl's at 18 months. Another reads a child's percentile and assumes it never changes. Both are rough, and both ignore the parents entirely, while the mid-parental formula ignores the child's own growth. That mismatch is why the calculator offers both estimates and blends them instead of choosing one. The table shows what each approach uses and what it leaves out.
| Rule of thumb | What it uses | What it ignores |
|---|---|---|
| Mid-parental formula | Mother's and father's heights, plus or minus 13 cm by sex | The child's own growth so far |
| Double the height at age 2 (boys) or 18 months (girls) | One early measurement | The parents, and the timing of puberty |
| Stay on the same percentile | The current percentile on a growth chart | Early or late puberty, which can shift it |
| Growth-based predictor | Current age and height against a reference table | Family patterns beyond what the table assumes |
How to get taller: what actually helps, and what does not
Genetics sets the range; environment decides how close to the top of it a person lands. Four factors matter while growth plates are still open:
- Nutrition. Adequate protein, calories, calcium, vitamin D and zinc during childhood and adolescence support normal bone growth. This is not just theory: average population height has risen measurably across generations in many countries as nutrition and childhood health improved, a well-documented pattern called the secular trend in height. Chronic malnutrition or untreated illness, in the other direction, can permanently reduce final adult height.
- Exercise. Regular activity supports healthy bone density and normal growth-plate development, though there is no good evidence that exercise itself adds height beyond a person's genetic potential.
- Sleep. Growth hormone is released mainly during deep sleep, so consistent, adequate sleep through the growing years supports a normal growth trajectory.
- Posture. Good posture does not add height; it just stops slouching from making someone look shorter than they actually are, by straightening the spine's natural curve rather than lengthening any bone.
Once growth plates fuse after puberty, typically by the mid-teens for girls and the late teens for boys, height is fixed. No diet, exercise routine, stretching regimen or product changes that.
Measuring height properly before you predict anything
Every predictor is only as good as the measurement that goes into it. Measure barefoot on a hard, flat floor with heels, hips and shoulders against a wall, eyes looking straight ahead, and lay a flat book across the top of the head so it touches the wall. Measure at about the same time of day each time, because people are roughly a centimeter taller in the morning than in the evening as the spine compresses during the day. Take three readings and use the middle one. Children under two are measured lying down, which gives length rather than standing height. If you are asking how tall am I for a height test at home, this routine is what makes the number comparable with the one at the doctor's office.
Why the range matters more than the number
A predicted height is the middle of a spread, not a promise. The mid-parental estimate carries a range of about 8.5 cm either way, so a predicted 179 cm for a boy really means somewhere between roughly 170 and 188 cm, with most children landing near the middle. Two siblings from the same parents can sit at opposite ends of that range, which is ordinary genetic variation rather than a sign that something is wrong. Where the range becomes useful is at the edges. A child whose measured height is tracking far above or below the predicted band, or who falls across several percentile lines in a short time, is worth mentioning to a pediatrician, who can check growth rate, puberty timing and general health. A calculator can flag that the numbers look unusual; it cannot explain why.
The honest limits of any height prediction
Both methods on this site, and every method described above, are built on population averages. A "late bloomer" who starts puberty later than typical peers often simply reaches their eventual height later, not a shorter one, which is exactly the kind of individual variation a population-average formula cannot see in a single reading. Treat any predicted number here as a starting point for curiosity, not a diagnosis, and bring a genuine growth concern, unusually fast, slow, or off-pattern for a family's own history, to a pediatrician who can track it properly over time.