Abstract Study question Could the recall of toxic products have been avoided if the samples had been tested with a more sensitive Mouse Embryo Assay (MEA)? Summary answer Standardizing MEA protocols with sensitive methodologies could prevent the potential harm imparted on human embryos and avoid the need to recall initially unidentified toxic products. What is known already Dozens of consumables are used during an IVF cycle, with a potential effect on gametes/embryos in culture. Bioassays, such as the MEA, are commonly used to pinpoint embryo-toxic products/batches before their release. No strict regulations exist on how to perform these tests, although several variables have been described to play a major role on the assay’s sensitivity and its capacity to detect embryo-toxicity. Toxic products/batches that are not identified may need to be recalled. However, recalls are typically issued after the products/batches are suspected of hindering embryo development in clinical laboratories, with the loss of irreplaceable biological samples. Study design, size, duration Only in 2023, four important batch recalls have been issued by regulatory bodies or manufacturers, concerning such diverse products as oil, media, capillaries, or follicle aspiration needles (FAN). Reportedly, thousands of couples undergoing IVF cycles may have been affected by this undetected toxicity in samples that were already MEA-tested by the manufacturers or external testing laboratories, following an undisclosed methodology. Eight of these lots (2xmedia, 3xcapillaries, 2xFAN) were retested with a standardized sensitive MEA protocol. Participants/materials, setting, methods The samples were retested following a high-sensitivity and previously published protocol, using 1-cell stage embryos from a sensitive hybrid mouse strain (B6CBAF1), supplementing with 5mg/ml HSA, and maintaining a low embryo density per droplet. A total of 462 embryos were cultured, allocated to the different test groups (21 + 42 embryos for Medium 1 and 2; 63 + 84+147 embryos for Capillaries 1, 2 and 3; 42 + 42 embryos for FANs 1 and 2, respectively). Main results and the role of chance Embryos were cultured at 37.3oC, 6%CO2 and 7%O2 for 5 days (96h). The expanded blastocyst formation rate (BFR) in each group was assessed as the end-point of the assays, and an 80% was considered as the toxicity-marking threshold. When cultured in Medium 1 and Medium 2, supplemented with 5mg/ml HSA, only 0/21 (0%) and 7/63 (11.1%) of embryos reached the expanded blastocyst stage, respectively. Capillaries and FANs were tested with an indirect method, in which the samples were first incubated with a small volume of control culture medium, and the extract was used to culture the embryos. With this approach, Capillaries 1, 2 and 3 yielded a 23.8% (15/63), 22.6% (19/84) and 0.7% (1/147) BFR, respectively. Similarly, when testing FANs 1 and 2, only 66.7% (28/42) and 19.1% (8/42) of the embryos were able to form a blastocyst. Unfortunately, the recalled oil batches were not available for retest. In parallel to each test, a control group was cultured with standard culture medium without contact with any of the samples; the BFR in all control groups was above 80%. These surprisingly low results suggest that toxicity was present in these samples in high concentrations, evidencing the need for their recall. Limitations, reasons for caution The methodology used in this project successfully identified all the recalled samples as embryo-toxic. However, the same protocol might not prove equally sensitive when testing other products, as the sensitivity of the MEA has been shown to be highly dependent on numerous external factors and to vary between different toxins. Wider implications of the findings Recent events could lead to the belief that the MEA is an insensitive bioassay to assess embryo-toxicity. Far from true, attention should be paid to the specific methodology used in each case. The lack of regulations and standardization translates into huge differences in sensitivity, even leading to completely contrary results. Trial registration number not applicable