Pregnancy Due Date Calculator

What Is a Pregnancy Due Date?

The due date is the anchor every other number in prenatal care hangs from — when screening is offered, how growth is judged, and what counts as early or late. It is worth knowing exactly what it is measured from before you decide how much weight to put on it.

The 280-Day Convention

Obstetrics counts a pregnancy as 40 weeks — 280 days — from the first day of the last menstrual period to the estimated date of delivery. ACOG's Committee Opinion 579 states that average directly, and Committee Opinion 700, reaffirmed in 2025, makes it the working rule: by convention, the EDD is 280 days after the first day of the LMP.

Two assumptions are built into that single number. The StatPearls chapter on pregnancy dating spells them out: the technique assumes a normal 28-day menstrual cycle, and that ovulation happens on day 14 of it. Everything this calculator does beyond plain addition exists because one or both of those assumptions is usually wrong for a given person.

Split the 280 days at the assumed ovulation point and the rest of the arithmetic on this page falls out of it: 14 days from the period to ovulation, then 266 days from conception to birth. Every method below is a different way of finding that same 266-day finish line from a date you can actually name.

Why the Clock Starts Two Weeks Before Conception

Counting from the last period means the first two weeks of an official pregnancy contain no pregnancy at all. It survives as the standard for one blunt reason: the first day of bleeding is an event you can observe and write down, and the moment of conception almost never is.

The practical consequence is a permanent two-week offset between the two ways of counting. Gestational age, the figure this calculator reports and the one every scan, screening window and term definition uses, is measured from the LMP. Fetal or embryonic age is measured from conception and runs about 14 days behind it — so 8 weeks pregnant describes an embryo roughly 6 weeks old.

The conception and IVF methods here do not escape that convention, they convert into it. Whichever start date you give it, the calculator works out an LMP-equivalent date 280 days before the due date and reports your weeks from there, which is why a day-5 transfer already reads as 2 weeks 5 days pregnant on the morning of the transfer itself.

If you are working forward rather than backward and want the fertile window that produced these dates, that is the job of the Ovulation Calculator.

What the Word "Estimated" Is Doing There

The panel closes every result with the same sentence: "Only about 1 in 20 babies arrives exactly on the due date — full term spans weeks 37–42". The first half is a rule of thumb rather than a measured constant; the second half matches the NHS, which puts it as pregnancy normally lasting from 37 weeks to 42 weeks from the first day of your last period.

There is a measured version of the same caution. Nguyen and colleagues checked predicted delivery dates against a Copenhagen series of 17,450 spontaneous singleton births; in the 14,805 of them with a reliable last-period date, the average gap between the predicted day and the actual one was 8.63 days, against 7.96 days for dates predicted from an ultrasound measurement of the fetal skull. Both methods are, on average, more than a week out.

So the useful output of this page is not one square on a calendar. It is a five-week landing zone with a labeled midpoint, and the arithmetic below is what fixes the midpoint as precisely as a date can be fixed.

How Do You Calculate a Pregnancy Due Date?

Every date quoted from here down was produced by running this page's own formula, or by counting days from what it returned, so the examples and the readout above cannot disagree. The running example throughout is a last period beginning Saturday, August 1, 2026.

The Formula

Four starting points, four constants. Pick the row that matches the date you know:

LMP:          due = first day of last period + 280 days + (cycle length − 28)
Conception:   due = conception date + 266 days
IVF day-5:    due = transfer date + 261 days
IVF day-3:    due = transfer date + 263 days

The three lower constants are the 280 days with the unknown stretches removed. Conception dating drops the 14 pre-ovulation days: 280 − 14 = 266. An IVF transfer drops those 14 days and the days the embryo has already been alive in the lab, so a day-5 blastocyst is 280 − 14 − 5 = 261 and a day-3 embryo is 280 − 14 − 3 = 263. Those two figures are not this site's invention — ACOG gives them in the same wording: for a day-5 embryo the EDD would be 261 days from the embryo replacement date, and for a day-3 embryo, 263 days.

The cycle-length term on the first line is the correction for ACOG's own stated caveat about the 280-day rule — that it does not account for irregularities in cycle length or variability in the timing of ovulation. It assumes the whole difference between your cycle and a 28-day cycle sits before ovulation, which is the standard assumption, not a measured fact for any individual.

Using the Calculator, Step by Step

The form has three fields and two buttons:

• Calculate From — four buttons rather than a dropdown: last menstrual period (LMP), conception date, IVF day-5 transfer, IVF day-3 transfer. LMP is selected when the page opens.
• Date — a date picker pre-filled with today's date. What it means changes with the method above it: the first day of bleeding, the day of conception, or the day of the embryo transfer.
• Average Cycle Length (LMP method) — a whole number of days from one period's first day to the next period's first day. The formula accepts 15 through 60 and rounds decimals to the nearest day, so 28.4 is treated as 28 and 28.5 as 29 — which is a one-day difference in the answer: Saturday, May 8, 2027 against Sunday, May 9, 2027 for an August 1 last period.
• Calculate runs the formula; Reset clears the panel back to its empty state.

Two behaviors are worth knowing before you use it. The page does not recalculate as you type — it runs when you press Calculate. It would have run once on its own had every field arrived filled, but the cycle-length box starts empty, so the panel opens blank rather than with a worked example already in it. Pressing Calculate with that box left empty still computes on the conception and IVF methods, where the formula never reads cycle length; only the LMP method refuses it.

The cycle-length box also carries no browser-level minimum or maximum, and the form is set to skip native validation, so nothing stops you typing 5 or 500. The 15-to-60 guard lives in the formula, and it answers with an amber notice rather than a date.

Worked Example: Last Period August 1, 28-Day Cycle

With the LMP method selected, August 1, 2026 in the date field and 28 in the cycle box:

cycle adjustment = 28 − 28      = 0 days
days to add      = 280 + 0      = 280 days
due date         = Aug 1, 2026 + 280 days

Hero readout

Saturday, May 8, 2027

Under the hero, the first breakdown line is fixed text — "40 weeks (280 days) from the LMP-equivalent start date" — and the second is the live one. It reports your gestational age today, the trimester it falls in, and the countdown, in this shape: "Today: 2 weeks 2 days pregnant (first trimester), 264 days to go". That example is the line the calculator produced for this input on August 17, 2026; the numbers in it move by one day every day, while the due date above them does not.

A conception date of August 15, 2026 returns Saturday, May 8, 2027 as well, which is the internal consistency check worth doing yourself: August 1 plus a 14-day wait to ovulation is August 15, and 266 days from there is the same Saturday.

Worked Example: The Same Date With a 35-Day Cycle

Keep August 1, 2026 and change the cycle length to 35. Nothing about the last period has changed; the estimate of when ovulation happened has.

cycle adjustment = 35 − 28      = +7 days
days to add      = 280 + 7      = 287 days
due date         = Aug 1, 2026 + 287 days

Hero readout

Saturday, May 15, 2027

One week later, and the whole schedule moves with it. The 20-week point shifts from December 19, 2026 to December 26, 2026, and the 41-week mark — the point at which the NHS says induction will be offered if labor has not started — shifts from May 15, 2027 to May 22, 2027. A tool that ignores cycle length would have handed this person the May 8 date and a week of misplaced appointments.

The correction runs both ways. A 24-day cycle from the same August 1 start returns Tuesday, May 4, 2027, and a 21-day cycle returns Saturday, May 1, 2027 — a full week earlier than the 28-day answer.

Worked Example: An IVF Transfer

IVF removes the two guesses that make LMP dating soft, because the embryo's age and the transfer date are both known to the day. Switch the method and the cycle-length box stops mattering.

The last column is not a coincidence and it does not depend on the calendar. A day-5 transfer is always 19 days into the LMP-equivalent count on the day it happens (14 days to the ovulation-equivalent plus 5 days of embryo age), which the panel prints as 2 weeks 5 days pregnant. A day-3 transfer is always 17 days, and a conception date always 14 — which the panel prints without a day count, as a plain "2 weeks pregnant".

Run the arithmetic backward and the four methods converge, which is the strongest evidence the page is internally consistent: a last period on August 1 with a 28-day cycle, a conception on August 15, a day-3 transfer on August 18 and a day-5 transfer on August 20 all return Saturday, May 8, 2027.

If what you actually want is the week count rather than the date at the end of it, that is the same math with a different output — see the Gestational Age Calculator.

Naegele's Rule, and Why It Drifts

The pencil-and-paper shortcut is Naegele's rule, attributed to the German obstetrician Franz Naegele (1778–1851): add nine months and seven days to the first day of the last menstrual period. Because calendar months are not equal, that is not the same operation as adding 280 days.

Across all 365 possible start dates in 2026, the two methods agree on 211 of them. On 62 the nine-months-plus-seven-days answer lands one day later than the 280-day count, on 61 it lands two days later, and on 31 it lands three days later. A last period on May 15, 2026 is a three-day case: this calculator returns Friday, February 19, 2027, while Naegele's rule gives February 22.

In all 365 cases the drift runs one way — the rule's date is never earlier than the counted one, and never more than three days later — and it is entirely an artifact of February and the 30-day months. This calculator adds real days, which is what ACOG's 280-day definition specifies. Lawson's 2021 review of the rule reached the sensible conclusion about the whole exercise: treat the result as a guideline for the expected date, not a definite date.

What the Calculator Refuses

Two inputs return the amber "check your inputs" notice instead of a date, and nothing is silently substituted for either:

• An empty or unreadable date field returns "Enter a valid date".
• On the LMP method, a blank cycle-length box, or one holding a number that rounds outside 15 to 60 days, returns "Enter a cycle length between 15 and 60 days". The rounding happens before the check, so the band actually accepted is 14.5 up to just under 60.5: 14.5 is read as 15 and answers Sunday, April 25, 2027 for an August 1 last period, while 14.4 is refused. A blank box does not quietly default to 28 — a guessed 28 for someone with 35-day cycles would move the due date a week.

The bounds are wide on purpose rather than clinical: outside 15 to 60 days the assumption that the difference from a 28-day cycle sits in the pre-ovulation phase stops describing a real menstrual cycle at all.

Pregnancy Due Date Chart

Four reference tables, all built from this page's own formula, so a cell here and the readout above will always agree. Every one assumes a 28-day cycle except the second, which is the cycle-length table itself.

Due Date by the First Day of Your Last Period

Find the month your last period started, then the column closest to the day it began. Each column is exactly seven days from the one beside it, so you can count forward or back from the nearest one.

Most rows follow the folk rule — same date, nine months on, plus a week — but the December, March, April, May, June and July rows do not, and that is the Naegele drift above showing up in a table. A period starting December 1, 2026 is due September 7, 2027, not September 8; one starting March 1, 2027 is due December 6, not December 8.

Every date here assumes a 28-day cycle. If yours is longer or shorter, move the answer by the difference: a 31-day cycle puts the date three days later than the cell, a 26-day cycle two days earlier.

Due Date by Cycle Length

The same last period — August 1, 2026 — read across every cycle length the calculator accepts. The middle column is what the formula actually adds.

From one last period, the accepted range of cycle lengths spans 45 days of due dates, from April 25 to June 9, 2027. That is the size of the error a calculator makes when it assumes everyone ovulates on day 14 — and it is why the cycle-length box is worth filling in honestly rather than rounding to 28 because 28 is the number on the box.

The correction is linear and small in the normal range: one day of due date per day of cycle. Inside the 24-to-38-day band that covers most adults, the whole spread is 14 days.

Milestone Dates for One Due Date

Fix the running example — last period August 1, 2026, 28-day cycle, due Saturday, May 8, 2027 — and the rest of the calendar follows from it. Gestational ages are written the way maternity notes write them, weeks and days.

Shift the whole table by however many days your own start date differs from August 1, 2026. The intervals never change, because they are all fixed counts of days from the same LMP-equivalent start.

Note how much of the calendar sits after the due date. From 40w0d to 42w0d is another two weeks, and pregnancies that reach it are still inside the range the NHS describes as normal.

Reading the Chart Backward

Given a due date, the same four constants run in reverse — useful when a scan has already assigned a date and you want to know which of your own dates it implies.

The LMP-equivalent column is the number the calculator itself works with internally: whichever method you choose, it subtracts 280 days from the due date to find a notional last-period date, and counts your weeks from there.

How to Read Your Result

The panel returns a date and up to three supporting lines. Each answers a different question, and one of them changes every night at midnight.

The Lines the Panel Returns

Using the August 1, 2026 example with a 28-day cycle:

The gestational-age line disappears whenever the LMP-equivalent start date has not arrived yet, and the panel comes back with three lines instead of four: the due date, the fixed convention line and the caveat. A future date does it, obviously. So does a recent past date on a long cycle, because the count runs from your date plus (cycle minus 28) rather than from the date itself — a period three days ago on a 35-day cycle puts the equivalent start four days from now, so there is nothing yet to count. The same happens once the count passes 315 days — 45 weeks, five weeks past the due date — which is far enough outside any plausible pregnancy that the calculator stops reporting a week number rather than printing an absurd one.

Weeks, Days and Which Trimester

Gestational age is the elapsed days from the LMP-equivalent start divided by seven: whole weeks, then leftover days. The trimester boundaries the calculator applies are ACOG's, from the same committee opinion that sets the 280-day rule — the first trimester runs up to and including 13 6/7 weeks, the second from 14 0/7 to 27 6/7 weeks, and the third from 28 0/7 weeks onward.

Written that way the three blocks are exactly equal: fourteen weeks each, 42 weeks in all. The third trimester has no formal upper limit — it is cut at 41w6d here so the symmetry is visible, and because 42 0/7 weeks is where the postterm label begins. Note where the due date sits inside it. At 40w0d you are twelve weeks into that final fourteen-week stretch, with two more weeks of normal pregnancy on the far side of it.

One quirk of the wording: leftover days are printed only when there are some. A count of exactly 14 days reads "2 weeks pregnant" rather than "2 weeks 0 days pregnant", and handing the calculator today's date as the last period produces "0 weeks pregnant".

Early Term, Full Term, Late Term, Postterm

The panel's closing line uses "full term" in its everyday sense, matching the NHS description of pregnancy normally lasting from 37 to 42 weeks. Clinically the label is narrower. In 2013, in a committee opinion since reaffirmed in 2025, ACOG and the Society for Maternal-Fetal Medicine endorsed replacing the single word "term" with four labels, because outcomes are not uniform across that five-week stretch.

The reason for the split is in ACOG's own summary: the frequency of adverse neonatal outcomes is lowest among uncomplicated pregnancies delivered between 39 0/7 and 40 6/7 weeks of gestation. The due date is not the edge of that window, it is inside it.

At the other end, the NHS position is that induction will be offered if labor has not started naturally by 41 weeks, and that anyone continuing past 42 weeks should be offered increased monitoring. In the running example those two dates are May 15 and May 22, 2027 — worth knowing in advance, because they are the conversations that happen if the date passes quietly.

Planning Against a Range Instead of a Day

The practical translation of everything above: build plans around weeks 37 to 42, not around the hero date. In the running example that is April 17 to May 22, 2027 — the window in which a normal birth is expected to happen, all five weeks of it.

That reframing changes real decisions. Leave that begins on the due date begins three weeks after the early-term window opened — April 17 against a due date of May 8 in the running example — and travel booked for 39 weeks is booked inside the full-term block rather than before it. The date to plan against for anything that must be finished beforehand is 37 weeks, not 40.

The other number that runs on this same week count is weight gain, which is tracked against gestational age rather than against the calendar — the Pregnancy Weight Gain Calculator.

Limits: When This Date Does Not Apply

This calculator can only be as good as two things: the date you give it and the assumption that ovulation followed it in the usual way. Each limit below is one of those two failing, and none of them is a reason to distrust the arithmetic — they are reasons to treat its output as a midpoint that a scan may move.

Half of Remembered Period Dates Are Not Accurate

ACOG's committee opinion puts the weakness of LMP dating in one sentence: it has been reported that approximately one half of women accurately recall their LMP. The date you type is a memory, and this calculator has no way to test it.

The scale of the resulting correction has been measured. In a study ACOG cites, 40 percent of women randomized to a first-trimester ultrasound had their due date adjusted because ultrasound and LMP dating disagreed by more than five days. In the control group, scanned in the second trimester, only 10 percent were adjusted — which ACOG reads as evidence that an early scan improves the estimate even when the first day of the last period is known.

A shifted or half-remembered start date moves everything downstream by the same number of days, silently and without any sign in the output that anything is wrong.

When an Ultrasound Overrides This Date

ACOG, the American Institute of Ultrasound in Medicine and the Society for Maternal–Fetal Medicine agree on the ranking: ultrasound measurement of the embryo or fetus in the first trimester, up to and including 13 6/7 weeks, is the most accurate method to establish or confirm gestational age. The thresholds below are where a scan stops confirming a calculated date and replaces it.

Put numbers on it with the running example. A last period of August 1, 2026 makes October 24, 2026 exactly 12w0d, so a scan that day is compared against that. A scan measuring 11w2d implies a due date of May 13, 2027 — five days from the calculated May 8, inside the threshold, so the original date stands. A scan measuring 10w6d implies May 16, eight days out, and the due date is changed to match the scan.

Once a date is set this way it is meant to stay set: ACOG says subsequent changes to the estimated due date should be reserved for rare circumstances, discussed with the patient, and documented. A later scan measuring big or small is a growth finding, not a new due date. ACOG also asks for confirmation early — a pregnancy with no scan confirming or revising the date before 22 0/7 weeks is considered suboptimally dated, which in the running example means before January 2, 2027.

The Cycle Correction Is a Patch, Not a Fix

Adjusting for cycle length repairs the most obvious flaw in the 280-day rule and leaves the deeper one alone. It assumes that every day your cycle differs from 28 is a day added to the follicular phase before ovulation, and that the interval from ovulation to birth is otherwise fixed.

The second half of that assumption is the weak one, and it has been quantified. In the NIEHS Early Pregnancy Study, where ovulation was identified from urinary hormone measurements rather than assumed, Jukic and colleagues found the coefficient of variation of gestational length was 4.9 percent when measured from the last menstrual period against 3.7 percent when measured from ovulation. The gap between those two figures is the variability in ovulation timing — real, measurable, and not something an average cycle length can remove.

Which is also why an average is the wrong input for a cycle that swings. Running the same August 1 start at a 24-day cycle and a 35-day cycle returns May 4 and May 15, 2027 — an eleven-day span with no single answer inside it. Someone in that position gets more from an early scan than from a more carefully chosen average.

Real Pregnancy Length Varies by More Than a Month

The strongest evidence against over-reading any due date comes from the small number of pregnancies where the start point was known exactly. Jukic and colleagues followed 125 naturally conceived singleton pregnancies with ovulation identified hormonally, and reported a median time from ovulation to birth of 268 days — 38 weeks 2 days — with the range of gestational length spanning 37 days even after six preterm births were excluded.

Two things follow. First, this calculator's conception method adds 266 days, two days short of that cohort's median, and its 280-day LMP rule carries the same small offset. Nguyen's Copenhagen data pointed the same way: adding 282 days instead of 280 reduced the average error from 8.63 to 8.41 days and cut the share of births classified as post-term from 7.9 percent to 5.2 percent, while raising the share classified as preterm from 3.96 to 4.48 percent. Their conclusion was that an LMP-only estimate should use 282 days.

This page stays with 280 because that is the convention every maternity service, screening window and set of notes runs on, and a due date that disagrees with your own hospital's by two days is worse than useless. But it is worth knowing the round number is a convention, not a measurement — Lawson's review found ethnicity, height, cycle variation, ovulation timing, parity and maternal weight all significantly influence how long a pregnancy runs.

Where the Arithmetic Stops Describing Anything

In these situations the last-period method is not approximate, it is inapplicable, and the answer above should be treated as a placeholder until a scan replaces it:

• Cycles that are genuinely irregular, or conditions such as PCOS and thyroid disease where there is no stable cycle length to enter.
• A pregnancy conceived soon after stopping hormonal contraception, when the first cycle off it may not have behaved like a typical one.
• A pregnancy conceived while breastfeeding, where ovulation can return before any period does — there may be no last period to count from at all.
• Bleeding in early pregnancy mistaken for a period, which starts the count from the wrong event and shifts the due date by roughly a month.
• Twins and higher-order multiples: the dating arithmetic is identical, but what is planned around the date is not, and that is a conversation with a maternity team rather than a calculator.

In every one of these, the first-trimester scan is not a second opinion — it is the primary method, and the calculator is the placeholder.

If your cycles are regular but you have never counted one, the average you need is the gap between two period start dates — which is the first thing you get out of the Period Calculator.

This Page Is Not Prenatal Care

A due date calculator does arithmetic on a date you supply. It cannot confirm a pregnancy, see a scan, know your history, or notice that something is wrong, and no output here should delay a call to a midwife, doctor or maternity unit.

Use it for what it is good for: getting oriented before a first appointment, converting an IVF transfer date into a week count, checking that the date on your notes matches the date on your calendar, and planning around a five-week window instead of a single square.