Preimplantation genetic testing (PGT) is highly accurate, but not perfect. The specific accuracy depends on which type of PGT you’re getting and what it’s looking for. When testing for chromosome abnormalities (PGT-A), the positive predictive value is around 96%, meaning roughly 4 in 100 embryos flagged as abnormal may actually be viable. When testing for a known single-gene disorder like cystic fibrosis or sickle cell disease (PGT-M), the misdiagnosis rate drops below 0.1%. These numbers are reassuring, but the biology of early embryos introduces complications that make accuracy more nuanced than a single percentage suggests.
The Three Types of PGT
PGT isn’t one test. It’s a family of three, each looking for something different. PGT-A screens for aneuploidy, which means having too many or too few chromosomes. This is the most commonly discussed type and the one with the most debate around its value. PGT-M tests for a specific inherited condition that runs in your family, like Huntington’s disease or BRCA mutations. PGT-SR looks for structural rearrangements in chromosomes, such as translocations, where chunks of chromosomes swap places.
PGT-M is the most precise of the three, with misdiagnosis occurring in fewer than 1 in 1,000 cases. PGT-SR, which often uses the same sequencing technology as PGT-A, shows similarly strong concordance rates. PGT-A is the one where accuracy gets complicated, because the biology it’s trying to assess is inherently messy.
How Accurate PGT-A Really Is
When a large multicentre evaluation compared PGT-A results against the gold standard of amniocentesis, newborn blood tests, and pregnancy outcomes, the concordance rate was 99.8% for embryos called normal (euploid). Out of 507 confirmed outcomes, only one embryo had a discrepancy: PGT-A called it normal, but amniocentesis later revealed trisomy 21. That’s an extremely low false-negative rate for chromosomally normal calls.
The trickier side is false positives, where PGT-A labels an embryo abnormal when it might have been capable of producing a healthy pregnancy. One study using a type of advanced genetic analysis found the positive predictive value of PGT-A was 96%. A separate study transferred 102 embryos that PGT-A had called aneuploid (abnormal), and none resulted in an ongoing pregnancy, putting the clinical error rate somewhere between 0% and 2.43%. So while a small percentage of “abnormal” results may be wrong, the test is clearly identifying embryos with real chromosomal problems in the vast majority of cases.
Why Mosaicism Complicates Results
The biggest biological limitation of PGT-A is mosaicism. Early embryos are dynamic, rapidly dividing organisms, and not every cell in an embryo necessarily has the same chromosomal makeup. Some cells might be normal while others carry an extra or missing chromosome. In clinical data, roughly 12% of embryos show mosaic patterns.
Here’s why that matters for accuracy: the biopsy takes a small sample of cells from the trophectoderm, which is the outer layer that will become the placenta, not the inner cell mass that becomes the baby. If the sampled cells happen to be abnormal but the inner cells are normal, the test produces a false positive. The reverse is also possible but harder to detect. Embryos can also self-correct as they develop, pushing abnormal cells to the outer layer and keeping normal cells in the core. This means some embryos labeled as abnormal or mosaic could theoretically have developed into healthy pregnancies.
Extrapolating from a handful of randomly selected outer cells to predict the fate of the entire embryo is, as researchers have noted, “bound to result in errors.” Overinterpreting abnormal findings can lead to discarding potentially viable embryos, which reduces the overall probability of pregnancy from a given IVF cycle.
What ASRM Says About PGT-A
The American Society for Reproductive Medicine released a committee opinion in 2024 that may surprise people who assume PGT-A is standard care. Their position: “The value of PGT-A as a routine screening test for all patients undergoing in vitro fertilization has not been demonstrated.” They also stated that routine blastocyst biopsy with aneuploidy testing for all infertile patients cannot be recommended at this time.
This doesn’t mean PGT-A is useless. It means the evidence doesn’t support using it for everyone. Recent multicenter randomized controlled trials found that overall pregnancy outcomes with frozen embryo transfer were similar between PGT-A tested and untested groups. The potential benefit of reducing miscarriage risk also remains unclear based on current studies. For specific patient populations, particularly those with recurrent pregnancy loss or repeated implantation failure, the calculus may be different, but the blanket recommendation isn’t there.
PGT-A and Maternal Age
One of the most common reasons clinics recommend PGT-A is advanced maternal age, since older eggs produce more chromosomally abnormal embryos. The test does reliably identify this: when euploid embryos are transferred, pregnancy rates per transfer are similar across all age groups. A 42-year-old transferring a chromosomally normal embryo has roughly the same chance of getting pregnant as a 33-year-old transferring one.
The catch is getting to that point. As age increases, the chance of producing a blastocyst suitable for biopsy drops, and the chance of that embryo being euploid drops further. The probability of achieving a successful pregnancy per egg retrieval cycle falls from about 29% for patients under 35 to just 5% for patients over 40. PGT-A accurately identifies which embryos are normal, but it can’t create normal embryos when the egg quality isn’t there. For older patients with very few embryos, testing may eliminate their only options, including embryos that might have self-corrected.
The Biopsy Itself
Beyond the test’s diagnostic accuracy, there’s a small physical risk from the biopsy procedure. A skilled embryologist removes 5 to 10 cells from the trophectoderm at the blastocyst stage (day 5 or 6 of development). The risk of damaging the embryo during this process is less than 1% to 2% in experienced hands. This is a significant improvement over older techniques that biopsied at the cleavage stage (day 3), when the embryo had far fewer cells and removing even one represented a larger proportion of the whole.
PGT-M: Testing for Specific Genetic Conditions
If you’re testing embryos for a known single-gene disorder, the accuracy picture is much clearer. PGT-M has a misdiagnosis rate below 0.1%, making it one of the most reliable genetic tests available. This is partly because the test is customized for your specific mutation, with a probe designed to detect exactly the genetic variant in question. There’s less ambiguity than with PGT-A because you’re looking for a defined, binary answer: does this embryo carry the mutation or not?
For couples who carry genes for serious conditions like Tay-Sachs, spinal muscular atrophy, or hereditary cancer syndromes, PGT-M is a powerful and well-validated tool. Prenatal testing (amniocentesis or chorionic villus sampling) is still typically recommended to confirm results during pregnancy, but discordant findings are exceedingly rare.
What This Means for Your Decision
PGT is accurate, but “accurate” means different things depending on the test. PGT-M for a known genetic condition is extremely reliable, with error rates below 1 in 1,000. PGT-A for chromosome screening is technically proficient, with concordance rates above 99% for normal results and a positive predictive value around 96%, but biological realities like mosaicism mean it occasionally flags viable embryos as abnormal. The test is better at confirming that a normal embryo is truly normal than it is at guaranteeing that every abnormal result is truly hopeless.
Your age, the number of embryos you produce, and why you’re doing IVF all factor into whether PGT-A adds meaningful value to your cycle. For someone with many blastocysts, it can help prioritize transfers and potentially reduce time to pregnancy. For someone with very few embryos, it may eliminate options that had a small but real chance of success. PGT-M, on the other hand, is broadly recommended whenever a known heritable condition is at stake.