Study question We sought to explore the levels of myeloid-derived suppressor cells (MDSCs) in women experiencing recurrent miscarriage (RM) or recurrent implantation failure (RIF) Summary answer MDSCs were significantly expanded in RM patients with respect to fertile healthy women (p = 0.03, 8.25 vs 7.00). What is known already RM and RIF are two distinct disorders that can be associated with a baseline pro-inflammatory state. Different surrogate inflammatory biomarkers have been associated with these disorders in a subgroup of patients, including expanded cytotoxic NK cells, increased pro-inflammatory cytokines or reduced regulatory T cells (TReg). MDSCs are a heterogeneous population of leukocytes, known for their capacity to modulate immune responses. The immunosuppressive function of MDSC in cancer has been well established, while a controversial role in autoimmune diseases has been reported. Typically, there are two major groups of MDSCs in humans: granulocytic/polymorphonuclear MDSCs (PMN-MDSCs) and monocytic MDSCs (M-MDSCs). Study design, size, duration A cross-sectional study to evaluate the levels of MDSCs in women with RM and RIF consecutively referred to our Reproductive Immunology Unit. 55 patients and 23 healthy women were evaluated from 2021 to 2022. RM was defined as the loss of two or more pregnancies and not necessarily consecutive. RIF was defined as the failure to achieve a clinical pregnancy after transferring at least four good-quality embryos. Participants/materials, setting, methods A full fertility screening was performed for all women and their partners as part of routine care. A group of healthy non-pregnant women up to 45 years old with two or more children (proven fertility) and no spontaneous miscarriage was used as control group. Freshly collected blood samples were analyzed by multiparametric flow cytometry using a BD FACS CANTO II. Peripheral blood M-MDSCs were characterized based on HLA-DR, CD14, CD33 and CD45 expression. Main results and the role of chance In this study, we evaluated 55 women, 35 that fulfilled RM criteria and 20 with RIF. In addition, a group of 23 non-pregnant fertile women was included as a control group. We characterized MDSCs as CD14+CD33+CD45+HLA-DR-/LOW, according to previous reports. We observed a significant expansion of peripheral blood MDSCs in RM group compared to the control group (8.25% [6.62-11.9] vs 7.05% [6-7.8], p = 0.03) and RIF group showed a similar increasing trend compared to controls (8.6% [5.65-12.35] vs 7.05% [6-7.8], p = 0.09). We performed ROC curves analysis to evaluate the predictive power. We established an optimal cut-off level of 8.3% in RM group, with a sensitivity of 50% and a specificity of 86.96%. Moreover, women with more than 8.3% had a higher probability of presenting RM compared to the control group (OR: 6.66; CI: 1.68-26.46). MDSCs are characterized by their ability to exert a potent immunosuppressive function. However, recent investigations have shown that MDSCs could act as pro-inflammatory cells under certain inflammatory conditions, such as autoimmune diseases. In this context, MDSCs could worsen the pro-inflammatory state in women, leading to miscarriage or implantation failure. Limitations, reasons for caution In order to validate these results, we should extend our cohort and control group. Moreover, we need to perform a functional evaluation of this subset so we could elucidate their role in these conditions. Wider implications of the findings MDSCs might be a good predictor for RM and RIF patients and with others (as expanded NK cells) may classify a subgroup of patients with a pro-inflammatory profile that can benefit from personalized therapies. Trial registration number Not aplicable