You need three things to predict your adult height at home: your parents' measured heights, your own current height and weight, and your age. With those you can run every method below in about five minutes. Here they are, ranked from least to most accurate, with a worked example for each.
One rule before you start: measure, do not guess. Parents overestimate their own height, and so do teens. Use a wall, a flat object on your head, and a tape, in the morning. Full measuring guide here.
Key Takeaways
- Mid-parental height takes 30 seconds and gives a genetic baseline, ±8.5 cm.
- Khamis-Roche adds your own height, weight, and age and is about twice as tight (±4 to 5 cm, 90% of the time).
- Growth-chart tracking uses your percentile over time and is the method most likely to catch a late or early growth spurt.
- Bone-age X-ray is the only method that can see how much growing is physically left, and it needs a doctor.
Method 1: Mid-Parental Height (30 seconds, ±8.5 cm)
This is the formula pediatric endocrinologists use as a first pass.
Step 1. Measure both parents in cm.
Step 2. Add them together.
Step 3. Boys add 13 cm, girls subtract 13 cm.
Step 4. Divide by 2.
Worked example (boy): Mother 165 cm, father 180 cm.
165 + 180 = 345. Add 13 = 358. Divide by 2 = 179 cm (about 5'10.5").
Worked example (girl): Same parents.
345 − 13 = 332. Divide by 2 = 166 cm (about 5'5.5").
Your range: about ±8.5 cm around that. So the boy above should expect somewhere between roughly 170 and 188 cm, most likely near 179.
In feet and inches: add the parents' heights in inches, add 5 inches for a boy or subtract 5 for a girl, divide by 2.
Why so wide? The formula only knows your parents. It cannot see whether you got the tall genes, the short ones, or the mix. Is height just genetics? covers what the rest of the picture looks like.
Method 2: Khamis-Roche (5 minutes, ±4 to 5 cm)
The most accurate published method that does not need an X-ray. It takes your age, sex, current height, current weight, and mid-parental height (from Method 1), then applies age-specific coefficients from a long-term growth study.
The coefficients change with every half-year of age, so the formula is not something to do by hand. Use a Khamis-Roche calculator, or the estimate inside TallerTeen, which starts from the same inputs. What matters is that you enter measured values:
1. Your height, morning, shoes off, against a wall.
2. Your weight, same day.
3. Both parents' heights, measured rather than remembered.
4. Your exact age in years and months.
What the output means: in the original validation, 90% of boys ended up within ±5.3 cm of the prediction, and 90% of girls within ±4.3 cm. Half of all predictions were within about 2 cm. That is as good as it gets without an X-ray.
Known blind spot: Khamis-Roche assumes an average puberty timeline. If you are a clear late bloomer (no growth spurt by 15 as a boy, or 13 as a girl) it tends to under-predict you, because it reads your current height as "already most of the way there." Late bloomers should mentally add to the top of their range. Read the full accuracy comparison.
Method 3: Growth-Chart Percentile Tracking (ongoing, the one that catches spurts)
This is what your pediatrician does at every checkup, and you can do it at home with the CDC growth charts.
Step 1. Find the CDC stature-for-age chart for your sex (2 to 20 years).
Step 2. Plot your height at your current age. Note the percentile line you are closest to (for example, 60th).
Step 3. Follow that same percentile line out to age 18 to 20. Where it ends is your projected adult height.
Step 4. Re-plot every 3 months. If you stay on the same line, your projection holds.
Worked example: A 14-year-old boy at 165 cm sits near the 50th percentile. The 50th percentile line ends around 176 to 177 cm at age 20, so that is his projection.
Why this method is special: it is the only home method that notices when you change lines. A teen who jumps from the 40th to the 60th percentile in six months is in a growth spurt, and the projection moves up. One who drifts from the 60th down to the 40th may have finished early, or may be sleeping or eating too little to keep up. Either way, that is information the one-shot formulas cannot give you.
Known blind spot: percentile tracking assumes you will stay on your line, which is exactly what late bloomers do not do. Before their spurt they drift down the chart, then jump back up. If you are a late bloomer, this method under-projects you until the spurt starts. Signs your growth plates are still open helps you tell which case you are.
Method 4: Bone-Age X-ray (needs a doctor, sees what is left)
Not a home method, but it is the one to know about. A doctor X-rays your left hand and wrist and compares the growth plates to a reference atlas to get your bone age. A 15-year-old with a bone age of 13 has about two extra years of growth left compared with a 15-year-old whose bone age is 15.
Combined with your current height, bone age gives the most direct estimate of remaining growth. Its published accuracy is still around 4 to 6 cm, and it over-predicts many late bloomers, but it answers the one question none of the home methods can: how much runway is actually left?
When to ask for one: if you are well below your mid-parental range, have crossed two percentile lines downward, or are 15+ with no signs of puberty. Those are the situations where a pediatrician would want the X-ray anyway.
Putting It Together: What to Actually Do
1. Run mid-parental height to get your genetic window.
2. Run Khamis-Roche (or the estimate in TallerTeen) to narrow it using your own numbers.
3. Track your percentile every 3 months so you see spurts and stalls as they happen.
4. Treat the result as a range, then work on the things that decide where in the range you land: 8 to 10 hours of sleep, enough protein and calcium, daily movement, and posture. How much sleep teenagers need to grow is the highest-leverage place to start.
Your genes set the window. Your habits pick the spot inside it.
FAQ
Can I predict my height without knowing my parents' heights?
Yes, with less precision. Percentile tracking (Method 3) does not need them. Khamis-Roche works with an estimated mid-parental height, but the error grows.
Does the "double your height at age 2" rule work for teens?
No. It is a rough rule for toddlers with no published error bounds, and it tells you nothing you cannot already see on a growth chart by your teens.
My prediction says I will be shorter than my dad. Is that possible?
Yes. Mid-parental height is an average, and you inherit a random mix of both parents' genes. Siblings of the same parents commonly differ by 10 cm.
I am 17. Is it too late for these methods to matter?
The predictions still work, but the range is small because most of your growth is done. Can you grow taller after 18? covers what is realistic at that point.
Sources: Khamis & Roche, Pediatrics 1994; NIH/PMC on self-reported parental height; CDC Clinical Growth Charts; NIH StatPearls on puberty and growth; AAP HealthyChildren on stages of puberty.