When can you find out the baby's gender?

The earliest reliable answer comes from a blood test at 10 weeks. The most common answer comes from the anatomy scan between 18 and 20 weeks. Almost everything offered in between is guesswork dressed up as science, and the reason is anatomical rather than technological.

Here is what each method can actually tell you, when it becomes reliable, and why a scan before 14 weeks is so often wrong.

NIPT blood test, from 10 weeks

Non-invasive prenatal testing works on a genuinely elegant principle. Small fragments of placental DNA circulate in the pregnant person's bloodstream, so a single blood draw contains fetal genetic material. The laboratory looks for sequences from the Y chromosome. Present means male, absent means female.

From 10 weeks onwards there is normally enough fetal DNA in the sample for this to be highly reliable, with reported accuracy above 99% when the fetal fraction is adequate. It is worth understanding what the test is really for, though: NIPT is a screening test for chromosomal conditions such as trisomy 21. Fetal sex is a by-product of the analysis, not its purpose.

A few situations can produce a misleading result. A vanishing twin can leave Y chromosome material behind. A low fetal fraction, more common earlier in pregnancy and at higher maternal weight, can return no result at all. Rarely, a maternal chromosomal variation interferes. These are uncommon, but they are the reason a laboratory reports a probability rather than a certainty.

The anatomy scan, 18 to 20 weeks

This is the mid-pregnancy ultrasound offered routinely almost everywhere, and it is where most people actually find out. Its real job is to examine the brain, heart, spine, kidneys and limbs. By this stage external genitalia are well formed and, given a clear view, identification is correct in the region of 99% of cases.

The qualification matters more than the number. Accuracy depends on getting a clear view, and that depends on how the baby is lying, how much amniotic fluid surrounds it, and how the ultrasound beam penetrates. A baby with legs crossed or a hand in the way is simply not going to be assessable that day, however experienced the sonographer.

Why scans before 14 weeks get it wrong

Early in development, male and female embryos look the same from the outside. Both start with an identical structure called the genital tubercle. Differentiation begins around 7 to 9 weeks under hormonal influence, but the external result stays visually ambiguous for several weeks after that.

By roughly 14 weeks the anatomy has diverged enough to be distinguishable on a good scan. Before that, an ultrasound is not being read incorrectly. There is genuinely very little to read.

This is also the honest answer to the "nub theory" scans marketed at 12 to 14 weeks, which judge the angle of the genital tubercle. Reported accuracy varies enormously between studies and between operators, it improves the closer you get to 14 weeks, and it is not something clinical services offer as a determination.

Every method, side by side

CVS and amniocentesis

Chorionic villus sampling, from about 10 to 13 weeks, and amniocentesis, from about 15 to 20 weeks, both sample fetal cells directly and produce a full chromosome analysis. That makes sex determination definitive rather than probabilistic.

Neither is ever performed to satisfy curiosity. Both are invasive procedures carrying a small risk of pregnancy loss, so they are offered when there is a medical reason to examine the chromosomes, usually a screening result or a family history. If you are having one, sex information comes along with the result you were actually seeking.

After IVF with genetic testing

Where embryos have undergone preimplantation genetic testing, the chromosome analysis already includes the sex chromosomes. The information exists before the embryo is ever transferred.

Whether the clinic shares it is a separate question, and depends on the law where you are being treated as much as on clinic policy.

The methods that do not work

The Chinese gender chart performs at roughly chance, which is exactly what you would expect from a table that maps lunar age against conception month. It is fun, and it is right half the time for the same reason a coin is.

The same applies to fetal heart rate, the shape of the bump, skull theory, Ramzi theory, and whether you are craving sweet or salty things. None has evidence behind it. Heart rate in particular gets repeated confidently and has been examined and found not to differ meaningfully between male and female fetuses.

Sex, gender, and a note on the law

Worth stating plainly, because the words get used interchangeably: what any of these tests identify is fetal sex, determined by chromosomes and visible anatomy. Gender is a separate thing, and not something a scan or a blood test can report.

One practical point for readers outside the UK, US and Europe. Several countries, including India and China, legally restrict the disclosure of fetal sex, as a measure against sex-selective abortion. If you are being treated in one of those jurisdictions, your clinician is not permitted to tell you regardless of what the scan shows.

Frequently asked questions