Abstract Study question Are dysregulated levels of immunomodulators (CXCL14, IL-15, and TIMP3) associated with aberrant decidual responses in the human endometrium? Summary answer Low transcript levels of CXCL14, IL-15, and TIMP-3 were associated with uterine natural killer (uNK) cell deficiency and pro-senescent response in the endometrium. What is known already In the mid-luteal phase, the endometrium becomes receptive to embryo implantation, while simultaneously poised for menstruation. Both events require decidualization, the differentiation of endometrial stromal cells into two cell subpopulations: progesterone-dependent decidual cells (DC) and progesterone-independent senescent decidual cells (snDC). Homeostatic balance between DC and snDC dictates endometrial fate. In non-conception cycles, snDC predominate as progesterone levels fall leading to menstruation. In conception cycles, endometrial transformation requires recruitment and activation of uNK cells, orchestrated by DC release of potent immunomodulators, CXCL14, IL-15, and TIMP3. This activation enables uNK cells to eliminate snDC, transforming the endometrium into the decidua of pregnancy. Study design, size, duration Endometrial biopsies (n = 730) were obtained from women attending the Implantation clinic at University Hospitals Coventry and Warwickshire (UHCW) NHS Trust. Collection was timed across the peri-implantation window based on the pre-ovulatory LH surge (LH + 6-11 days). Each biopsy was appropriately processed for immunohistochemistry and RT-qPCR analysis. Participants/materials, setting, methods Transcript levels of immunomodulators (CXCL14, IL-15, and TIMP3) were quantified in whole tissue by RT-qPCR, and uNK cells by immunohistochemical staining of CD56+. Due to changing expression levels across the luteal phase, results were normalized to the day of cycle using percentile charts. Expression levels of immunomodulators and uNK cells were compared to the expression of biomarkers of DC and snDC populations (SCARA5 and DIO2 respectively). Main results and the role of chance To understand the relationship between immunomodulators and decidual cell homeostasis, patient samples were separated into distinct categories of putative endometrial defects on the basis of DC and snDC biomarker expression. These included those with pro-decidual response (n = 61), pro-senescent response (n = 100), uNK cell deficiency (n = 42), and opposingly, uNK cell excess (n = 54). Expression of all tested immunomodulators (CXCL14, IL-15, and TIMP3) was significantly reduced in samples with a pro-senescent response (high DIO2, low SCARA5), compared to those displaying a pro-decidual response (high SCARA5, low DIO2; P < 0.0001; Kruskal-Wallis test). Additionally, patients with high numbers of uNK cells had a ∼2-fold-increase in CXCL14 (P = 0.0023), and IL-15 (P = 0.0052), and a 3-fold increase in TIMP3 (P < 0.0001) expression, compared to those with uNK cell deficiency. The data suggests that abnormal levels of these immunomodulators may dysregulate uNK cell activity and disrupt endometrial homeostasis, ultimately leading to miscarriage. Limitations, reasons for caution In this study, we showed the relationship between immunomodulators and decidual subpopulations. At present, it is not clear whether dysregulated levels of immunomodulators cause endometrial dyshomeostasis or are a result of the impaired decidual response. Further investigation is required to explore the “chicken-or-egg” dilemma. Wider implications of the findings Analysing these immunomodulators in a large number of samples provides a deeper understanding of the homeostatic relationship between uNK cells, DC, and snDC during the implantation window. This may be an invaluable diagnostic tool to identify women at risk of miscarriage, and in monitoring the effectiveness of pre-pregnancy interventions. Trial registration number N/A