My Baby ETA

How the Calculators Are Built

The models, sources, and calibration choices behind each My Baby ETA calculator.

Last reviewed: September 7, 2026

This page describes each calculator's model in enough detail to check it. Every number below is either taken from a named source or marked as a calibration choice. The calculator pages repeat the parts that matter for reading a result; this page is the complete account.

Labor timing calculator

What it estimates. The chance that labor begins on its own on each remaining day of the pregnancy. "On its own" means regular contractions starting or the membranes rupturing before a planned induction or cesarean. Scheduled and emergency deliveries are outside the model.

Structure. The forecast is a probability distribution over gestational days from 20 weeks through 43 weeks 2 days, built from two parts. The before-37-weeks part carries a baseline 5% of the total, rounded from the 4.5% spontaneous preterm rate among U.S. singleton births reported by Gyamfi-Bannerman and Ananth (2014), and is spread across preterm weeks in the proportions summarized by Goldenberg and colleagues (2008). The from-37-weeks part is a normal curve truncated to the term window, with its center chosen so that the whole distribution's median lands on day 284 for a first birth and day 282 for a later birth, the medians Smith (2001) reported for non-elective deliveries.

Calibration. The term curve's spread, 9.44 days, is not a reported value. It was chosen so that the median targets above are met and the curve also reproduces a late-term benchmark: Marquette and colleagues (2014) found that 67.6% of uncomplicated pregnancies still ongoing at 41 weeks 0 days began labor spontaneously by 41 weeks 6 days.

Adjustments. Answers about a previous birth, previous preterm birth, IVF, smoking, age, BMI, and miscarriage history change the forecast by amounts taken from the studies listed in the sources, and in every case the applied amount is at or below the published estimate. Factors linked to preterm birth change only the before-37-weeks part. Previous birth timing, age, and BMI move the term curve by fractions of a day to about one day. The relationship between one pregnancy's length and the next is weak, with a correlation of 0.10 in Ng and Steer (2016), so a previous 41-week labor moves this forecast by days, not weeks.

Updating. Each calculation removes days that have already passed and rescales what remains, so the result always describes the forecast from today forward.

Deliberately excluded. Fetal sex, interpregnancy interval, race, cervical examination findings, and biomarkers. Reported associations for the first two are small or endpoint-dependent; the last two are not stable inputs a person can supply at home. Race is not treated as a biological timing variable.

The calculator page carries the formula-level statement of the same model, including the density tilting used for term adjustments.

Due date and pregnancy week calculator

What it does. Applies standard obstetric dating arithmetic to whatever dates you enter, shows one estimated due date per source, and, when you enter more than one source, compares them.

Rules. Last menstrual period plus 280 days, adjusted by the difference between your usual cycle length and 28 days. Known ovulation or fertilization plus 266 days. Embryo transfer date minus the embryo's age at transfer plus 266 days. An ultrasound report's gestational age is converted to a due date by calendar arithmetic. These follow the American College of Obstetricians and Gynecologists' Committee Opinion 700 on estimating the due date.

Reconciling sources. When dates disagree, the calculator selects one primary source by clinical priority and reports the others alongside it with their differences. It never averages dates. When candidates differ by more than seven days, it recommends discussing the discrepancy with your care team rather than choosing for you. An established clinician due date, when entered, is used unchanged and outranks everything else.

Conception timing. One conventional estimate is shown. Jukic and colleagues (2013), measuring ovulation directly in 125 naturally conceived singleton pregnancies, found gestation lengths from 247 to 284 days after excluding preterm births and related conditions, a 37-day span. Conception timing derived from a period date or ultrasound inherits that variation.

Deliberately excluded. Raw crown-rump length formulas, hCG-based dating, automatic redating decisions, and any attribution of a pregnancy to a particular intercourse event.

Child growth planner

What it does. Estimates when a child is likely to reach a clothing or car-seat height or weight limit by continuing their current growth position forward along a reference curve.

References. WHO 2006 length- and weight-for-age standards before age 2 and CDC 2000 stature- and weight-for-age references from age 2, the combination CDC recommends for U.S. practice. A child's measurement is converted to a z-score with the reference's LMS parameters, the z-score is held constant into the future, and the reference is inverted at each future age to produce a projected measurement. The crossing date is the first future age at which the projected value reaches the limit, found numerically. Corrected age is used through age 2 for children born before 37 weeks. Standing height entered before age 2 has 0.7 cm added for comparison with the recumbent-length reference.

Ranges. For infants whose crossing falls inside a published WHO growth-increment interval, the displayed range uses the WHO 5th, 50th, and 95th percentile increments. Otherwise the range holds lower and higher z-score channels and re-solves the crossing. These are scenario bounds, not calibrated prediction intervals.

Head clearance. Optional. From age 2, seated height reference curves were fitted to 6,012 NHANES III sitting-height measurements for ages 24 to 108 months using survey-weighted cubic median regression, separately by sex. The coefficients are printed on the calculator page. Below age 2, broad seated-growth shares of 35% to 65% of projected length gain are applied instead.

Limits. Clothing limits are the upper values from each brand's published chart. Car-seat limits are recorded per model, revision, and mode from manufacturer documentation, with a link to the source in each row.

Child height predictor

What it does. Estimates adult height for children aged 4.0 through 17.5 years using the corrected Khamis-Roche equation, which takes exact age, current height, current weight, and the average of both biological parents' heights.

Source and population. Khamis and Roche (1994), with the 1995 coefficient erratum. The equation was fitted to 223 boys and 210 girls in the Fels Longitudinal Study, all White and from southwest Ohio. It is the most widely used non-radiographic prediction method, and its cohort is narrow. The calculator states this on the result.

Error range. The paper reported average 90% absolute-error bounds of 2.10 inches for boys and 1.68 inches for girls across the supported ages. The calculator adds and subtracts the published bound for the child's sex. Error is larger during puberty, when timing varies most between children.

Comparisons. The current height percentile uses the CDC 2000 stature-for-age reference. The family range uses the conventional mid-parental formula, which adds 13 cm to the parents' average for boys and subtracts 13 cm for girls, then halves the sum, shown with a ±10.2 cm range. Both appear next to the primary estimate so the reader can see how much the methods agree.

Principles that apply across the site

Maintenance

Each model has an evidence register recording the sources, the implemented values, and the classification of each value as reported, adapted, or calibrated. Registers are re-read at least once a year and whenever a guideline a calculator relies on changes. Automated tests check the labor model's baseline preterm share, medians, late-term benchmark, and monotonic probabilities on every change to the code. Corrections received through the contact page are checked against the cited source before a change is made.

Sources

Labor timing

Pregnancy dating

Child growth and height