Weeks, days, and the fourteen-day offset
Obstetric weeks are counted from the first day of the last menstrual period and written as weeks plus days: 22w 3d means twenty-two full weeks and three days. Conception happens around two weeks into that count, so the number is always about a fortnight ahead of the age of the embryo. It looks like an error and it is not. Cycle start dates are known; fertilisation dates are not, and a convention that everyone shares beats a more literal one that nobody can measure.
The practical effect is that a positive test taken a week after a missed period lands at roughly five weeks, and the first appointment is often booked for a couple of weeks after that. By the time most people begin thinking in weeks at all, they are already into the second month of the count.
Two ways to run this, and which to prefer
Last-period mode applies Naegele's rule: last period plus 280 days. Due-date mode runs the same arithmetic backwards, deriving a working last-period date by subtracting 280 days from whatever date the clinic supplied. The second is the better input whenever it exists, because a due date issued after a first-trimester scan already carries the correction that period-based dating cannot make. Details of how the two are reconciled are on the due date calculator.
Windows that close
Most of the schedule is flexible by a week or two. A few items are not, and those are the reason to convert weeks into dates at all.
| Item | Window | Why the timing is fixed |
|---|---|---|
| Nuchal translucency | about 11w 0d to 13w 6d | The fluid collection being measured is only present and measurable in that band |
| Cell-free DNA screening | from about 10 weeks | Below that there is often too little fetal DNA in the sample to report a result |
| Anatomy ultrasound | 18 to 22 weeks | Earlier the structures are too small to assess; later the position makes views harder |
| Glucose screening | 24 to 28 weeks | Insulin resistance rises through pregnancy and this is where the test discriminates best |
| Group B strep culture | 36 to 37 weeks | Colonisation comes and goes, so a result much older than five weeks stops predicting the state at delivery |
Trimesters as a rough division
The three-way split is convention rather than physiology, and the boundaries are drawn slightly differently by different sources. The version used here — first through 13w 6d, second to 27w 6d, third from 28w — is the common one. What the divisions describe loosely is a first period dominated by organ formation and the highest rate of miscarriage, a middle stretch where most people feel physically best and where the main structural assessments happen, and a final stretch of growth, monitoring and preparation for delivery.
Nothing changes on the boundary days themselves. If a week number in this output prompts a question about whether something is normal, the answer is not in the table; it is with the person providing prenatal care, and it is worth asking early rather than watching a calendar.
Questions people ask
How many months pregnant does a given week make me?
Clinical practice does not use months, which is why the question has no clean answer. Dividing by four gives nine and a half notional months for a forty-week pregnancy, while calendar months make it a little over nine. Both conversions are approximate and neither is used for anything, since every screening window, growth reference and clinical decision point is written in weeks. If someone asks, dividing the week number by 4.3 gets close enough for conversation.
Why does the fetal length in the output jump between 18 and 20 weeks?
The measurement changes. Before about 20 weeks fetal length is reported crown to rump, because the legs are curled and cannot be measured reliably. After that it is reported crown to heel, which adds the whole lower body at once. The fetus does not grow six inches in a fortnight; the ruler moved. Ultrasound reports handle this the same way, so a length that appears to leap between two scans may simply be two different measurements.
Can I use this to work out conception date?
Only roughly, and it is worth being careful with the result. Subtracting fourteen days from the working last-period date gives an approximate conception date on the assumption of a 28-day cycle with mid-cycle ovulation, and both parts of that assumption vary. Sperm can remain viable for several days, so the fertile window spans about six days rather than one. Any conception date derived this way carries at least several days of uncertainty in each direction and should not be treated as precise.
The clinic says a different week than this page. Why?
Almost always because the pregnancy was redated after an early scan and the calculation here is still running from a period date. Crown-rump length in the first trimester is the more accurate clock, and once a practice adopts a scan-based due date it manages everything from that. Switch this calculator to due-date mode, enter the date the clinic gave, and the weeks will line up. Where a discrepancy of more than a day or two persists after doing that, the clinic record is the one to trust.