How Pregnancy Dating Actually Works

The two-week offset, stated plainly

Gestational age is the number of days since the first day of your last menstrual period. It is not the age of the embryo. In the 28-day cycle the convention assumes, ovulation happens around day 14, so for the first two weeks of a pregnancy measured this way there is no pregnancy at all — only a cycle that had not yet ovulated. The arithmetic follows directly. Six weeks of pregnancy is 42 days from the period, so the embryo is about 42 − 14 = 28 days old, four weeks. Eight weeks of pregnancy holds a six-week embryo. Twelve weeks holds a ten-week embryo. Forty weeks, at the very end, holds a thirty-eight-week one.

The convention survives because it is the only one that can be applied consistently. A date of conception is almost never known: intercourse and fertilisation are not the same event, sperm can wait several days, and ovulation itself is not directly observed. The first day of bleeding, by contrast, is a visible event most people can name. Some countries make the ambiguity explicit in the vocabulary. French practice counts semaines d'aménorrhée, weeks since the last period, and names the parallel count from conception semaines de grossesse — always two weeks fewer. Italian, Spanish and Portuguese practice have the same distinction. English and German clinical usage generally have only the menstrual count, which is why an English-speaking reader is more likely to be surprised by the offset than a French-speaking one.

Naegele's rule, derived and worked

The rule is: take the first day of the last period, add one year, subtract three months, add seven days. Adding a year and subtracting three months moves you forward by nine calendar months. Nine months averaged over a year is about 273 days, and the seven days on top bring the total to 280 — forty weeks. Worked on a real date: a last period on 5 January 2026 gives 5 January 2027, then 5 October 2026, then 12 October 2026. Counting the days between 5 January and 12 October gives exactly 280.

The calendar shortcut is not quite the same thing as the 280 days, and it is worth knowing which one your clinic uses. Because calendar months are unequal, the shortcut sometimes lands a day or three past 280. A last period on 30 March 2026 gives 6 January 2027 by the rule, which is 282 days, while 280 days lands on 4 January. Sweeping the rule over every possible last-period date in 2025 and 2026 — 730 dates — it returns exactly 280 days in 422 of them, about 58%, and 281, 282 or 283 days in the rest. The convention is the 280 days; the month arithmetic is a shortcut from before anyone carried a calculator. Every modern tool, and every dating wheel, counts the days.

What the rule assumes, and the correction for your cycle

Naegele's rule assumes a 28-day cycle with ovulation on day 14. If your cycle is longer, you ovulated later, conception happened later, and the due date should move later by the same amount. German obstetric teaching builds this in as the erweiterte Naegele-Regel: the due date is the last period plus 280 days plus the difference between your cycle length and 28. On a last period of 5 January 2026, a 32-day cycle gives 280 + 4 = 284 days, moving the date from 12 October to 16 October. A 25-day cycle gives 280 − 3 = 277 days, moving it to 9 October. The shift is exactly the difference in cycle length, no more and no less.

The assumption is worth questioning, because the 28-day cycle is a textbook average rather than a common experience. Bull and colleagues analysed 612,613 ovulatory cycles from 124,648 users in npj Digital Medicine in 2019 and found a mean cycle length of 29.3 days, with only about 13% of cycles lasting exactly 28 days. Fehring and colleagues, tracking 1,060 cycles in 141 women, found a mean of 28.9 days with 95% of cycles falling between 22 and 36 days, and more than 7 days of variation within the same woman in 42.5% of participants. If your own cycles vary that much, the cycle correction is an improvement on the raw rule but still an estimate — which is exactly why the scan exists.

When the scan wins: redating

In the first trimester every embryo grows at nearly the same rate, so a single measurement — the crown-rump length, from the top of the head to the bottom of the spine — pins the date tightly. ACOG's Committee Opinion 700 treats first-trimester ultrasound as the most accurate method available and says that when it disagrees with menstrual dating by more than a stated margin, the estimated due date should be changed to the scan's. The margin widens as pregnancy advances, because biological variation in size grows: more than 5 days up to 8 weeks 6 days, more than 7 days from 9 weeks to 15 weeks 6 days, more than 10 days to 21 weeks 6 days, more than 14 days to 27 weeks 6 days, and more than 21 days after that.

Practice differs a little between the six markets in when the scan is offered, though not in the principle. NICE guidance in the United Kingdom places the dating scan between 10 weeks 0 days and 13 weeks 6 days and switches from crown-rump length to head circumference once the length exceeds 84 mm, the same cut-off ACOG uses. French practice dates at the first-trimester scan between 11 and 13 weeks of amenorrhoea and, like ACOG, treats a discrepancy beyond about five days as grounds to take the scan's date. What matters for you is simpler than the guideline text: once your dates have been set by an early scan, that is the date the rest of your care will run on, and a later scan will not usually move it again.

Why IVF dating is exact

Everything uncertain about menstrual dating — when you ovulated, when fertilisation happened, how long the sperm waited — is known in a laboratory. ACOG 700 therefore says that when a pregnancy results from assisted reproduction, the assisted-reproduction date should be used and should not be overridden by a scan. The arithmetic starts from the same 280 days: subtract the 14 days from period to ovulation and you get 266 days from conception to birth. Subtract the age of the embryo at transfer and you have the number of days to add to the transfer date. A day-3 embryo gives 266 − 3 = 263 days; a day-5 blastocyst gives 266 − 5 = 261.

Worked on a date: a day-5 blastocyst transferred on 10 February 2026 gives a due date of 29 October 2026. The same answer falls out of the menstrual convention if you construct a virtual last period 19 days before the transfer — 14 days from period to ovulation plus the 5 days the embryo already had — which lands on 22 January 2026, and then add 280 days. Both routes give 29 October, which is the check that the two conventions are the same convention. It also explains why an IVF pregnancy is described in the same weeks as any other: the count is anchored to a menstrual period that never happened, which is odd but consistent.

What term actually spans

A due date is the midpoint of a distribution, not an appointment. ACOG's Committee Opinion 579 replaced the old single label of term with four: early term is 37 weeks 0 days to 38 weeks 6 days, full term is 39 weeks 0 days to 40 weeks 6 days, late term is 41 weeks 0 days to 41 weeks 6 days, and postterm begins at 42 weeks 0 days. In days from the last period those boundaries are 259, 273, 287 and 294. The due date at 280 days sits in the middle of the full-term fortnight, not at its edge, and the four labels exist because outcomes — respiratory ones in particular — differ measurably across a range that used to be described with one word.

The labels are American; the practice around them is national. French obstetrics speaks of a terme at 41 semaines d'aménorrhée and organises induction discussions around it, which is 287 days and the start of what ACOG calls late term. British, German, Italian, Spanish and Portuguese services each have their own thresholds for when a pregnancy continuing past the due date is monitored more closely or an induction is offered. None of that changes the arithmetic in this article, and all of it is a conversation to have with the team looking after you rather than a number to settle from a web page.

Frequently asked questions

Why is pregnancy counted from a period rather than from conception?

Because the date of a period is knowable and the date of conception usually is not. Ovulation is not directly observed, fertilisation can follow intercourse by up to five days because sperm survive that long in the reproductive tract, and Wilcox and colleagues showed in 1995 that conception can result from intercourse across a six-day window. The first day of bleeding is a single visible event, so it became the anchor. The price is the two-week offset: your gestational age is about two weeks more than the age of the embryo.

I know exactly which day I conceived — can I use that date instead?

You can convert it, and it is a reasonable estimate to bring to your first appointment: add 14 days to get a virtual last-period date, then add 280 days, which is the same as adding 266 days to the conception date. But the day of intercourse is not necessarily the day of fertilisation, so unless the pregnancy came from assisted reproduction the conversion carries its own error of a few days. ACOG's position is that first-trimester ultrasound remains the most accurate method available, so expect the scan to be the date your care runs on.

My due date changed at the scan. Which one is right?

The scan's, if it was done early and the difference exceeded the threshold for that gestational age — more than 5 days before 9 weeks, more than 7 days up to 15 weeks 6 days, and so on up the table above. That is not a correction of you; it means your ovulation did not fall on day 14 of your cycle, which is unremarkable. Once redated, the new date becomes the reference for every later decision, including monitoring past term, so it is worth knowing which date your notes carry.

Does a long cycle really push the due date later?

Yes, by the difference between your cycle length and 28 days, because a longer cycle almost always means a longer pre-ovulation phase rather than a longer post-ovulation one. A 32-day cycle moves the date 4 days later, a 25-day cycle moves it 3 days earlier. Bull and colleagues found the mean follicular phase to be 16.9 days against a mean luteal phase of 12.4 days, and it was the follicular phase that varied — a 95% range of 10 to 30 days against 7 to 17 for the luteal one. That is the mechanism behind the correction.

How likely is the baby to arrive on the due date itself?

The due date is one day inside a wide distribution, so landing on it is the exception rather than the rule. The useful framing is ACOG's: full term is a fourteen-day span from 39 weeks 0 days to 40 weeks 6 days, with early term for the two weeks before it and late term for the week after. Your due date sits in the middle of that fortnight. A birth several days either side of it is ordinary, and the labels exist to describe the range rather than to make one day special.