A due date is not a prediction of when a baby will arrive. It is a statistical anchor, and obstetricians build it from a small number of measurable inputs that vary in accuracy depending on how early in pregnancy they are taken. Understanding that hierarchy explains why a date can shift by a week after a single ultrasound, and why some parents are given two different dates at two different appointments.
Roughly 4 percent of babies arrive exactly on their estimated due date. About 60 percent arrive within a week of it, and around 90 percent within two weeks. Those numbers come from large cohort studies of spontaneous labor, and they are the reason clinicians talk about a “due window” rather than a due day. You can work out your own window with the pregnancy calculator, but the numbers behind it are worth understanding first.
The Naegele rule and the 280-day assumption
The oldest method still in daily use is the Naegele rule: take the first day of the last menstrual period, subtract three months, add seven days, and add a year. That yields 280 days, or 40 weeks. The rule assumes a 28-day cycle with ovulation on day 14, which is true for a minority of women.
If your cycle runs 35 days, ovulation likely happens around day 21, and the Naegele date will systematically land about a week early. Clinicians correct for this by adding or subtracting the difference between your cycle length and 28. A 32-day cycle moves the date back four days; a 24-day cycle moves it forward four. This is where “my doctor’s date is different from my app’s” usually originates.
Why ultrasound can override your period date
A first-trimester ultrasound measures crown-rump length, the distance from the top of the head to the bottom of the rump. Between 7 and 13 weeks, that measurement predicts gestational age with an accuracy of roughly plus or minus 5 to 7 days, and it does so consistently across different sonographers and different machines.
Because that precision beats the uncertainty of a recalled period date, the American College of Obstetricians and Gynecologists recommends that a first-trimester scan establish the due date when the two disagree by more than about 7 days. Later scans are less decisive — at 14 to 20 weeks the margin widens to roughly 7 to 10 days, and in the third trimester an ultrasound can be off by three weeks or more because babies of the same gestational age vary widely in size.
What actually changes the timeline
Several factors push a delivery earlier or later than the 40-week mark, and none of them are captured by arithmetic:
- First pregnancy: first-time mothers deliver on average about 5 to 7 days later than women who have given birth before.
- Multiple gestation: twins commonly deliver around 36 to 37 weeks, driven by uterine capacity rather than by the due date calculation.
- Maternal age: deliveries after 35 skew slightly earlier on average, though the effect is modest compared with pregnancy history.
- Ethnicity and stature: population-level differences of several days exist in gestational length, which is one reason a single universal formula is an approximation.
A worked example
Suppose the first day of your last period was March 3, and your cycles run 31 days. Naegele gives December 8. Correcting for the three-day cycle difference pushes that to December 11. At 9 weeks, a crown-rump length measurement corresponds to 9 weeks 2 days, which puts the ultrasound date at December 9.
The gap between the corrected period date and the scan is two days — smaller than the 7-day threshold — so the period-based date stands and the ultrasound simply confirms it. Had the scan instead indicated 10 weeks 1 day, the gap would exceed a week and the clinician would reset the due date to the ultrasound value.
What the 40-week figure really describes
Forty weeks is the median, not the mean, in many cohorts, and the distribution is left-skewed: more births cluster before 40 weeks than after. Gestational length in the largest studies ranges from about 37 to 42 weeks in uncomplicated pregnancies, with only a small fraction extending past 41.
This is why induction discussions typically begin around 41 weeks. The risk of stillbirth rises slowly but measurably after 41 weeks, and most guidelines recommend offering induction between 41 and 42 weeks rather than waiting indefinitely for spontaneous labor.
Practical takeaways for expectant parents
Treat the date as the centre of a two-week band. Plan leave and childcare around 38 to 41 weeks rather than a single day. Ask which method your clinician used to set the date — period-based, first-trimester scan, or a transfer from a fertility treatment — because that tells you how much confidence to place in it.
If you conceived through IVF, the date is unusually precise: it is calculated from the embryo transfer date and the age of the embryo at transfer, which removes both the cycle-length and ovulation-timing uncertainties entirely. If you are tracking symptoms and milestones week by week, the pregnancy calculator will convert a due date into gestational weeks so you can see exactly where you sit relative to the 40-week mark.
Frequently asked questions
Can my due date change more than once? It can, but good practice sets it once using the earliest reliable measurement. Later scans after about 20 weeks should not reset the date because fetal size variation by then is larger than the measurement error.
Is a 42-week pregnancy overdue? It is post-term by definition. Most providers will have offered induction before that point, and continuing past 42 weeks is uncommon in managed care settings.
Does the due date affect anything besides planning? Yes — it drives the timing of screening windows, growth assessments, and decisions about induction, which is why accuracy in the first trimester matters more than most people expect.


