Your age isn't the problem. Your embryo's chromosomes might be.

If you're over 35 and going through IVF, chances are your RE has shown you the chart. You know the one. Success rates drop off a cliff around your age and approach zero by the early forties. It's one of the first things they show you, and it's hard to unsee. SART's national data puts live birth rates per intended egg retrieval at 42.8% under 35, falling to 30.5% at 35-37, 19.4% at 38-40, 9.4% at 41-42, and 2.8% over 42. The CDC publishes similar figures and warns that aggregate rates may not reflect your individual outcomes.

That chart is real. But it's answering the wrong question.

A 2025 study of PGT-A cycles found no significant difference in pregnancy rates per euploid embryo transfer across age groups. The p-value was 0.61. A 42-year-old transferring a chromosomally normal embryo had statistically the same shot as a 32-year-old. The age curve isn't measuring whether your embryo will work. It's measuring how hard it is to find one that will.

That's a completely different problem. And it changes what you should actually be paying attention to.

Age doesn't make embryos worse. It makes them rarer.

The standard per-retrieval success rate mashes two separate forces into one number, and that's where the confusion starts.

The first force is aneuploidy. Older eggs produce more chromosomally abnormal embryos. At 30, roughly 30% of embryos carry the wrong number of chromosomes. By 37, it's closer to 40%. By 40, over 60%. By 43, it can hit 80% or higher. The chart drops steeply not because older embryos that pass chromosomal screening are somehow weaker, but because fewer of them pass at all.

The second force is egg count. Ovarian reserve declines with age. A 42-year-old might retrieve six eggs where a 30-year-old retrieves fifteen. Fewer eggs mean fewer embryos, which means fewer chances at a chromosomally normal one.

These two forces compound. You start with fewer eggs, and a higher fraction of the embryos those eggs produce are aneuploid. That's what the per-retrieval success rate is actually tracking. It's not measuring embryo quality. It's measuring attrition: how many eggs you start with, and what fraction survive the path from retrieval to chromosomally normal blastocyst.

Once you separate those two problems, the question changes. It stops being "am I too old for this to work?" and becomes "do I have a chromosomally normal embryo?"

Chromosomal screening is what actually moves the needle

Preimplantation genetic testing for aneuploidy (PGT-A) screens embryos for chromosomal normality before transfer. Whether it helps you depends on your starting point.

For patients 38 and older, the data is clear. A large SART analysis of 56,469 cycles found cumulative live birth rates of 51.3% with PGT-A versus 44.8% without for ages 38-40. A single-center study found 48.3% in the PGT-A group compared to 34.7% in untested transfers. For women with recurrent pregnancy loss, the benefit is even sharper: one study showed PGT-A nearly doubled live birth rates in patients under 38 who'd had three or more miscarriages (55.4% versus 32.0%).

For younger patients with lots of embryos, it's a different story. That same SART analysis found no improvement for patients under 35 (68.6% with testing versus 67.3% without). When most of your embryos are already chromosomally normal, screening doesn't add much. We've written a longer breakdown of how PGT-A works and who benefits most if you're weighing whether testing makes sense for your situation.

The per-transfer numbers tell the story most directly. Awadalla et al. found that euploid embryo transfers had live birth rates of 70.0% for autologous eggs in patients under 35 and 70.6% for donor eggs, versus 52.5% and 34.3% for untested transfers. When you know an embryo is chromosomally normal, success rates jump and the gap between age groups basically disappears.

Biopsy day matters too. Euploid embryos biopsied on day five had a 52.5% live birth rate, dropping to 43.2% on day six and 22.2% on day seven. The embryo's developmental pace carries information about implantation potential that euploidy status alone doesn't capture.

But euploid doesn't mean guaranteed. A 60-70% per-transfer success rate is far better than the alternative, but it isn't certainty. Uterine receptivity and endometrial factors play roles, alongside mechanisms we don't fully understand yet. Some patients transfer a beautifully graded euploid embryo and it doesn't implant. That's genuinely difficult. But it means the odds were 60-70%, not 100%.

And population-scale studies tracking outcomes across multiple complete cycles show that success is best understood cumulatively. A single euploid transfer gives you strong odds. Two or three give you very strong odds. The per-cycle framing that most clinics use makes every attempt feel like a coin flip. The cumulative framing is much closer to how this actually works.

A note on the research: the SART analysis draws from 56,469 cycles across hundreds of clinics, which gives it strong statistical power, but patients who chose PGT-A tended to be older and more likely to have experienced prior failures. The single-center studies control better for lab technique but are retrospective. True randomized trials are rare because once PGT-A became widely available, it became ethically difficult to assign patients to a "no testing" arm. These are real limitations. But the direction of the evidence is consistent across study designs, sample sizes, and patient populations: knowing your embryo is chromosomally normal is the single biggest variable you can control. It just isn't the only one.

The question after chromosomal screening

Here's where the question shifts.

Say you've done your retrieval and tested your embryos. You're looking at two or three euploid blastocysts. They're all chromosomally normal. They all have roughly similar odds of implanting.

But they're not identical.

Each of those embryos carries a different combination of the hundreds of common genetic variants that influence risk for conditions like type 2 diabetes, heart disease, Alzheimer's, and breast cancer. Two embryos from the same cycle are biologically siblings, and siblings can end up with meaningfully different genetic predispositions. One might carry substantially higher inherited risk for a condition that runs in your family. Another might not.

Standard PGT-A can't see any of this. It tells you whether the chromosome count is right. It doesn't tell you anything about which common variants each embryo inherited.

This is what polygenic screening addresses. It doesn't change the chromosomal picture. It adds a second layer of information: among your viable embryos, which ones carry lower inherited disease risk?

We built our polygenic scores specifically for this context. Moore, Davidson et al. tested 17 disease scores on sibling pairs, because embryos from the same cycle are siblings and that's the test that actually matters. Sixteen of seventeen scores showed no decrease in predictive performance within families. That's the validation standard that makes these scores meaningful in the one context where you'd actually use them: choosing between your own embryos. You can read the full technical paper for the complete methodology.

An independent study by Folkersen et al. confirmed the concept at scale, with over 40,000 sibling pairs. They found that families with five euploid embryos could see 27-67% relative risk reduction across nine diseases.

This isn't a replacement for PGT-A. It's what comes after it. PGT-A tells you whether an embryo is chromosomally viable. Polygenic screening tells you something different: among your viable embryos, which ones carry lower inherited disease risk? The two layers are complementary.

And the two tests don't have to mean two biopsies. We developed a technology called ImputePGTA that extracts polygenic risk information from the same data already collected during standard PGT-A testing. If you've already had PGT-A done through another lab, that data can often be used to generate polygenic scores without touching your embryos again.

The value is greatest for families with elevated baseline risk. If both parents have type 2 diabetes, or there's a strong family history of heart disease or Alzheimer's, the spread of genetic risk across your embryos tends to be wider, and the difference between the highest-risk and lowest-risk embryo is more meaningful. We incorporate family history into our risk estimates because it changes the baseline against which each embryo's score is interpreted.

One in 37 US babies is now conceived through IVF, and nearly 87% of cycles use frozen embryos. The overwhelming majority of IVF families already have a window where this kind of screening can happen.

The real question isn't "am I too old for IVF to work?" It's "what do I know about this embryo before I transfer it?" If you're thinking about what those answers look like for your family, our counselors can walk you through what the scores mean for your specific situation. You can reach out to us here.