How this instrument works
Birth weight percentile answers a different question than the postnatal growth charts pediatricians use after birth: instead of comparing a baby's weight to other babies of the same age since birth, it compares birth weight to other babies delivered at the same gestational age. A baby born at 27 weeks weighing 700 g and a baby born at full term weighing 3,400 g can both be entirely typical for their own gestational age, even though the numbers look nothing alike — this calculator answers 'is this weight typical for babies born this early (or this late)', which is the basis for small-for-gestational-age (SGA) and large-for-gestational-age (LGA) classifications.
This calculator uses the reference curves from Olsen et al. (2010, Pediatrics), built from a final analytic sample of 257,855 singleton U.S. infants born 22 to 42 weeks' gestation who survived to discharge, using the LMS method (Cole 1990): at each gestational week, three parameters — L (skewness), M (median), S (spread) — describe the birth-weight distribution for that sex. The entered birth weight is converted to a z-score against those parameters, then to a percentile using a logistic approximation to the normal curve, the same approximation already used by this site's IQ percentile calculator.
Two limits worth stating plainly. First, this calculator's reference table holds 10 representative gestational weeks (odd-week centers spanning roughly 23 to 41 weeks) from the Olsen 2010 published curves, not the full week-by-week resolution — it matches the entered gestational age to the nearest of those 10 points rather than interpolating, so a result for a week between two table points is computed as if it were exactly at the closer one. Second, this is a screening tool, not a diagnostic one: an SGA- or LGA-range result is a prompt for a clinician to review the birth alongside gestational dating accuracy, maternal history, and the baby's clinical exam — it is not, by itself, a diagnosis of growth restriction or any other condition.
- Choose Boy or Girl — the reference curves are sex-specific.
- Enter gestational age at birth in weeks (23 to 42).
- Enter birth weight in grams.
- Read the percentile, the z-score, and the reference median birth weight matched for that gestational age and sex.
- Bring a result near or beyond the 10th or 90th percentile to the baby's clinician for interpretation.
Worked example — a boy born at 31 weeks at the reference median
A boy born at 31 weeks weighing exactly 1,631 g: the Olsen 2010 reference median for a male infant at that gestational age is 1,631 g, so z = 0 — precisely the 50th percentile, the midpoint of the reference population by definition.
Compare a girl born at 27 weeks weighing 700 g. The reference median for a female infant at that gestational age is 936 g, so z works out to about −1.39, which the logistic approximation converts to roughly the 9th percentile — in the range clinicians watch for growth restriction, and exactly the kind of result that would prompt a closer look at prenatal growth scans and gestational dating rather than a conclusion from birth weight alone.
Now a boy born at 39 weeks weighing 4,000 g against a reference median of 3,469 g for that gestational age: z is about 1.03, landing around the 85th percentile — a larger-than-typical but not extreme birth weight for a term boy, a data point a clinician would weigh alongside maternal history (for example gestational diabetes) rather than in isolation.
Questions
What does a low or high birth weight percentile actually flag?
A percentile below the 10th is generally what defines small-for-gestational-age (SGA) and above the 90th generally defines large-for-gestational-age (LGA) in clinical use, though exact cutoffs and their significance depend on the reference used and the clinical context. Either flag is a prompt for a clinician to review the pregnancy history, gestational dating accuracy, and the baby's exam — it is not, by itself, a diagnosis of intrauterine growth restriction, macrosomia, or any other condition.
Why does this use a different reference than the CDC growth-percentile calculators on this site?
Birth weight is compared across babies born at different gestational ages, not across babies of the same age since birth the way postnatal weight-for-age charts work — a fundamentally different clinical question that needs its own reference curves. This calculator uses Olsen et al. (2010, Pediatrics), a large modern U.S. cohort built specifically for birth weight by gestational age, rather than the CDC 2000 postnatal charts this site's other growth calculators use.
Why does the calculator only use 10 reference weeks instead of every week?
Olsen et al. (2010) published LMS parameters across the full range of gestational weeks in their cohort, but this calculator hard-codes a representative 10-point subset (odd-week centers from roughly 23 to 41 weeks, taken verbatim from the published curves) and matches the entered gestational age to the nearest one — it does not interpolate between weeks or invent any in-between value. Results for a week between two table points reflect the closer reference week, so treat them as approximate.
How accurate is the percentile number itself?
The percentile is computed with a logistic function as a stand-in for the exact normal cumulative distribution function, which is the same well-documented approximation this site's IQ percentile calculator uses. It differs from the exact table lookup by roughly half a percentile point in most cases — close enough for a screening read, but not a substitute for a clinician's interpretation.
Can this calculator diagnose intrauterine growth restriction or macrosomia?
No. This is a screening and reference tool that performs a statistical comparison against a population curve — it cannot diagnose intrauterine growth restriction, macrosomia, or any other condition. Those diagnoses depend on a clinician reviewing prenatal growth trends, gestational dating, delivery circumstances, and a full newborn exam, not a single birth-weight percentile.
Does gestational age need to be exact?
Accuracy matters a lot here, since the reference curve shifts meaningfully week to week in preterm ranges — a gestational age that's off by even a week or two, especially estimated late or from an inaccurate last-menstrual-period date rather than an early ultrasound, can shift the percentile noticeably. Use the most clinically confirmed gestational age available, generally the one recorded in the delivery record.
References
- Olsen IE et al. — New Intrauterine Growth Curves Based on United States Data. Pediatrics. 2010;125(2):e214-e224
- CDC — Growth Charts overview: screening use and clinical interpretation
Read this first: This instrument computes a screening figure from population formulas — it is not a diagnosis, and it cannot see the whole picture a clinician can. Use it to inform a conversation, not to replace one.