Medical disclaimer. This article is general information, not medical advice. Only your healthcare provider can date your pregnancy, and an early ultrasound will beat any calculator. If a date matters legally or medically, ask for it to be documented in your notes.
If you are asking "when did I conceive?", the arithmetic is short: take your estimated due date and subtract 266 days. That lands on your most likely conception date. Most people reach this question backward — a scan or an app handed them a due date, and now they want to know which week of their life produced it. Reversing the calculation works. What it gives you is a six-day window rather than a single night, and the reason for that is the interesting part.
The math: your due date minus 266 days
Pregnancy runs on two different clocks, and almost every conception-date muddle comes from mixing them.
Gestational age starts on the first day of your last menstrual period. That is the clock your chart uses, the clock behind "you're 12 weeks along," and the clock the NHS means when it says pregnancy normally lasts 37 to 42 weeks from that first day. The estimated due date sits at 280 days — 40 weeks — on this clock.
Fetal age starts at fertilisation, about two weeks later. On this clock the same due date is 266 days away — 38 weeks.
So reverse-dating is just a subtraction between the two:
- Most likely conception date = due date − 266 days
- Dating LMP = due date − 280 days
- The 14 days between them is the assumed gap from the start of your period to ovulation.
If you know your last period instead of your due date, run it forward: conception ≈ LMP + (cycle length − 14) days. For a 28-day cycle that is 14 days after your period started. For a 34-day cycle it is 20 days — nearly a week later than the textbook answer, and the same six-day shift then moves your due date too. (The longer version of the forward math is in our guide to how due date calculators work.)
Conception means ovulation, not the night you remember
This is the single most useful thing to understand about the question, and one study settles it.
Wilcox, Weinberg and Baird followed 221 women charting daily urine hormones through 625 cycles. Conception occurred only when intercourse fell within a six-day period that ended on the estimated day of ovulation. The probability of conceiving ranged from about 0.10 five days before ovulation to about 0.33 on the day of ovulation itself — and no conceptions at all followed intercourse after that day.
Read that carefully and it tells you two things about the body. Sperm can survive long enough to fertilise an egg up to five days later. The egg's own fertile span is short enough that the window slams shut on ovulation day.
Two consequences follow for your date. First, "we only had sex once that month" narrows the fertile window but does not pin the conception date — fertilisation may have waited days. Second, a conception date derived from a due date describes ovulation, not an event on your calendar.
Which is why the honest answer is a window
A well-built reverse calculator returns two things: the most likely conception date, and a six-day window ending on it. That window is not padding for safety. It is the Wilcox window — the span in which conception is biologically possible, oriented so that ovulation sits at the closing edge, because nothing conceives after it.
A tool that hands back one date and no range is not more precise. It is just not showing you how wide the estimate is.
Implantation is a separate event — and a later one
People often reach for the date they got a positive test. That is neither conception nor implantation.
After fertilisation the embryo travels and then implants. In the same cohort, among the pregnancies that lasted at least six weeks, the first appearance of the pregnancy hormone hCG came 6 to 12 days after ovulation, and 118 of those 141 — 84% — implanted on day 8, 9 or 10 (Wilcox and colleagues, NEJM 1999). Only after implantation does hCG rise enough for a test to see it. In a later analysis of the same women, 10% of clinical pregnancies had not implanted by the first day of the missed period, which puts the best possible sensitivity of an hCG test that day at about 90%, rising to about 97% a week later — and the authors add that real test kits, being less sensitive than the research assay, will do worse than that ceiling. So "I found out on the 20th" places conception near the 6th, give or take.
Implantation bleeding is a weaker clue than its reputation suggests. When the same team tracked daily bleeding diaries through early pregnancy, spotting rarely fell on the day of implantation and they reported no support for the hypothesis that implantation produces vaginal bleeding. It is not a date to build on.
Your cycle length is the assumption that breaks
The 266-day shortcut quietly assumes a 28-day cycle with ovulation on day 14. Real cycles are messier than that.
In a prospective study of 696 cycles, Wilcox, Dunson and Baird found the fertile window landing across a very wide span of cycle days — in only about 30% of women was the whole fertile window inside the classic days 10 to 17. Even women who regarded their own cycles as regular had a 1–6% chance of being in their fertile window on the day their next period was due.
So if your cycles run 32 or 35 days, shift your conception estimate later by the difference. If they run 25 days, shift it earlier. And if they are irregular, treat the calculated date as a centre point with real uncertainty around it.
One more wrinkle: if your due date came from an ultrasound, reverse-dating gives you the scan's opinion, not independent evidence. A dating scan measures the fetus and derives an EDD; running that EDD back through the 266-day formula simply returns what the scan implied. That is fine — the NHS offers that scan at around 10 to 14 weeks to work out how many weeks pregnant you are and set the due date, and ACOG treats an early scan as the thing that confirms or revises an EDD. Just do not treat the reversed date as a second, confirming source.
IVF is the exception: the date is known, not estimated
If you conceived through IVF, none of this cycle guesswork applies. The clinic knows the day of fertilisation and the embryo's age at transfer, so the due date is counted from the transfer date rather than reconstructed from a period. On the standard convention that puts conception at 2 weeks 0 days, a day-5 blastocyst transfer means you are 2 weeks and 5 days pregnant on the day of transfer. The due date calculator on this site has a transfer-date mode for exactly that, alongside last period and conception date.
So how accurate is any of it?
Worth holding both ends of the estimate loosely. Jukic and colleagues followed naturally conceived pregnancies dated from ovulation itself — the most precise dating available outside a fertility clinic — and found a median of 268 days from ovulation to birth. Even after excluding preterm births and pregnancy-related medical conditions, gestation ranged from 247 to 284 days: a 37-day spread.
Read that in reverse and it is the same message: a due date is an estimate, so a conception date recovered from it inherits that estimate's uncertainty and adds the cycle assumption on top. A six-day window is a fair answer. A single certain date is not.
When the exact date really matters
Three situations come up often, and all three deserve a real conversation rather than a calculator:
- An exposure before you knew. A medication, a drink, an X-ray, a vaccine. What your provider needs is the window and your cycle history, not a single date.
- Leave and paperwork. Employers and insurers work from the documented EDD in your notes, not from a reverse calculation.
- Questions about timing and paternity. A reverse-dated window commonly spans a week or more, and cycle variation widens it further. It cannot answer that question; DNA testing can.
In all three, the strongest evidence is an early ultrasound. ACOG treats a pregnancy with no scan confirming or revising the EDD before 22 weeks as suboptimally dated, and once that date is in your notes it is meant to stay there — changing it later is reserved for rare circumstances.
How ProbaBaby handles it
In the ProbaBaby app, the Conception Date Calculator works in both directions. From a due date it takes 280 days back to the dating LMP and 266 days back to the most likely conception date, then draws the six-day Wilcox window ending on that date rather than a single day. From a last period it applies the same cycle-length shift to both the conception date and the due date, so the two figures on screen never contradict each other. Its cycle input runs from 20 to 45 days and defaults to 28, so starting from a last period, a long or short cycle moves the answer instead of being ignored. Starting from a due date it assumes 28 days — an EDD carries nothing to recover your cycle length from. The screen names what it is built on: Wilcox 1995, ACOG Committee Opinion 700 and FIGO. It also picks a lane and stays in it — conception at LMP plus 14, the obstetric calendar — which lands a day apart from the ovulation tools, because those count the first day of your period as cycle day 1.
The app's Due Date Calculator is a separate tool that accepts a last period, a conception date or an ultrasound measurement. The web calculator on this site is built independently: last period, conception date, or an IVF transfer date. If you are still tracking cycles rather than a pregnancy, the app's fertility planning tools — ovulation, fertile window, implantation — run on the same shared conventions: ovulation at cycle length minus 14, and a six-day fertile window ending on it.