Abstract Study question If a clinic changes the genetic laboratory it works with, should it expect its new PGT-A results to be equivalent to those it received previously? Summary answer Changing the genetic service provider can result in a significant difference in the proportion of embryos classified euploid, with important clinical implications. What is known already Preimplantation genetic testing for aneuploidy (PGT-A) aims to distinguish potentially viable euploid embryos from embryos harbouring lethal chromosome abnormalities. Typically, IVF clinics perform biopsy at the blastocyst stage and send the resulting trophectoderm specimens to specialist genetic laboratories for analysis. However, many commercially available genetic methods have not been subjected to rigorous validation, leading to uncertainty about their accuracy and predictive value. It would be concerning if the classification of embryos, derived from the same patient population, and from the same clinic, varied depending on the company that provides PGT-A services. Such differences would be unlikely to reflect biological reality. Study design, size, duration Our clinic recently began working with a new genetics company. Our previous PGT-A reference laboratory issued results for 1,136 blastocyst biopsy specimens over ∼15 months. Subsequently, the second company tested 784 biopsy samples in 7 months. Patient populations during the two periods were essentially identical (average maternal age 38.0 years). Participants/materials, setting, methods The first company used whole genome amplification followed by next generation sequencing (NGS) to evaluate the relative amount of DNA from each chromosome. The second company’s method was highly validated, involving targeted amplification of thousands of sites in the genome, again followed by NGS and analysis of relative DNA quantity from individual chromosomes. Additionally, they genotyped numerous polymorphisms, which assist in the detection of haploidy/triploidy, and also reveals problems that can compromise accuracy (e.g. contamination). Main results and the role of chance The first PGT-A company classified 44.4% of blastocysts euploid, 4.1% low-level mosaic, 51.5% aneuploid. In contrast, the second company reported significantly fewer embryos aneuploid (44.0%; P = 0.0019). Even if the mosaics reported by the first company are considered for transfer, the second company is still associated with more potentially transferable embryos (relative increase greater than one-sixth). If PGT-A results from the second company are correct, it implies that potentially viable embryos may have been misclassified by the first company, risking their exclusion and loss of the pregnancies they might have produced. Conversely, if the first company is correct, the second company may be failing to detect some aneuploidies, leading to inadvertent transfer of abnormal embryos, lower pregnancy rates and a higher incidence of miscarriage. 299 transfers following PGT-A using the first company produced 171 pregnancies (74.7%), while 64 single embryo transfers have taken place after PGT-A with the second company, resulting in 53 pregnancies (82.8% per transfer). Losses (biochemical or miscarriage) affected 21.6% and 16.9% of pregnancies after PGT-A conducted by the first and second companies, respectively (ongoing pregnancy rates of 58.5% and 68.8% per transfer, respectively) Thus, there is no evidence that aneuploidies are being missed by the second company. Limitations, reasons for caution Although there were no obvious differences between the patients with embryos tested by the two companies, this study was not prospective nor randomized, so a possibility of undetected differences remains. The study was not sufficiently powered to test the significance of apparent differences in implantation, miscarriage and ongoing pregnancy rates. Wider implications of the findings It is important that IVF clinics appreciate that individual PGT-A methods differ significantly, and that some have undergone much more rigorous validation than others. It is recommended that clinics ask PGT providers to share their validation data, especially clinical studies confirming that embryos classified ‘aneuploid’ or ‘abnormal’ are truly non-viable. Trial registration number Not applicable